Full text of "Acta Physiologica Scandinavica"

See other formats

ACTA

PHYSIOLOGICA

SCANDINAVICA

VOL. 11

REDACTORES

F. Reenpaa a. Krogh E. Langfeldt

Helsinki Kjobcnhavn Oslo

G. Liljestrand (KDiTom

StoclihoUn

COLLABORANTES

G, AnLORES (Lund), Y. Airila (Helsinki), E. L. Backman (Uppsala),

G. Blix (Uppsala), J. Bock (Kjobcnhavn), R. Eoe (KjSbcnhnvn),

H. V. Eoler (Stockholm), U. S. v. Eoeer (Stockholm), A. FCleino
(Oslo), R. Qraktt (Stockholm), G. GOtblin (Uppsala), E. Hammarsten
(Stockholm), E. Hausen (Kjobcnhavn), K. Hansen (Oslo), E. Honwu-
CanrsTENSEN (Stockholm), G. Kahlson (Lund), P. Leeoaard (Oslo),
J. Lehmann (Goteborg), J. Lindhard (Kjobcnhavn), E. Lenesoaabd
(Kjobcnhavn), K. MSllee (Kjobcnhavn), R. Nicolaysen (Oslo), S. Oe-
SKOv (Aarhus), A. V. Sahlstedt (Stockholm), P. SohOnheyber (Aarhus),
P.E. SiMOLA (Helsinki),!. Teorell (Uppsala), H.Tueorell (Stockholm),
T, Thhnberb (Lund), A. Westereund (Uppsala), A. I. Virtanen

(Helsinki), Y. Zotterman (Stockholm)

STOCKHOLM 194G

Reprinted with the permission of Acta Physiologica Scandinavica
KaroUnska Instilurer, Stockholm

JOHNSON REPRINT CORPORATION Johnson Reprint Company Limited
1 1 1 Fifth Avenue, New York, N.Y. 10003 Berkeley Square House, London, W.l

Fm! rcpriniinf. 1964, Johnson Kcprin! Corporaiion

I’rintcd in the United Sintc; of America

YOL. 11. INDEX.

Ease. 1. (28. I. 1040.)

rag.

Determination of Inulin in Urine and Plasma. By PouL Kruhof- ^

Inulin as an’indicator for the Extracellular Space. By Pout Kbu-

The Significance of Diffusion and Convection for tlic Distribution

of Solutes in the Interstitial Space. By Foul Kruhoffeb . . 3T
On Acute Effects of Cigarette Smoking on Oxygen Consump-
tion, Pulse Rate, Breathing Rate and Blood Pressure in Working

Organisms. By Ants Juurup and Leonid Muido . : 48

Amino Acids and Related Compounds in the Haemolymph of
Oryctes Nasicornis and Melolontha- Vulgaris. By Hans H.

UssiNG 61

The Influence of Anoxia on the Gastric HCl-Secrction. By K.
Habtula and M. Kabvonen 85

Ease. 2—3. (27. IV. 194C.)

Microdetermination of pH in Saliva. By Bodie SoHmDT-NiEESEN 97
The pH in Parotid and Mandibular Saliva. By Bodil Schmidt-

Nielsen 104

Perception of Weight and the Phenomenal Regression to tho
Real Weight (Thing Constancy Phenomenon). By Eeva Ja-

LAVISTO I * Ill

On the Synthesis of Creatine in the Animal Body, By Gdnnar

Steensholt 131

On Methylation Processes in Etiolated Wheat Germs. By (Iunnar

Steensholt 133

Choline Esterases in some Marine Invertebrates. By Klas-Bebtil

Augdstinsson 141

Investigations of the Phosphatase Activity in Serum and Organs
after Ligation of the Common Bile Duct in Doga. By Inger Gad 151
Studies on Serum Phosphatase Activity in Relation to Experi-
mental Biliary Obstruction in Rabbits. 11. By Jorgen Hoff’-
MEYER, Olaf Jallinq and Fritz Schonheyder 160

vot.. H. INI>5;.\.

l’«K.

Til'- I’ro-'‘iu-c of ;i Sul^taiicc with Symiiathin E Projicrties in

Splicn Extract.'. I5y U. S. von* Eiruilt lOS

Trot, in .^f.'tal^.)li.^nl of Ti.viuc Ccll.s in Vitro. -!. By T.xct; A.STUCf

anil .Vt.iinnT Fi,-;cin;R 187

Acrol'ic Bccovery after Anacrohio'-is in Be.st nnd Work. By Euuxa

187

Interaction Between riltrinoijen ainil PoIy.Kacchnriile Polysulfuric

.\ciil«. By AsTittm and Jokck.v Pii-kp. 211

Til'- Influence of ft-.Strophantin on the Meclianical Proj)ertic.s of

Cardiac Muscle. By Gl’.v.var Lr.vniN 221

Hfferent Iinpul.-os in tlie .Splanchnic Nerve. By B. Gkrnan'DT, G.

IjIUk.stkanh and Y. Zottkilmak 230

'rise ICffect of I{e.s{)iratory Cluinees npon the Spontaneous Injury
Di'idiarce of Afferent Mamninlian nnd Human Nerve Fibres.

By B. GKr.N'ANKT and V. Zottruman' 21S

Method.' for Continuou.s Ti.ssuc Culture a.s Aj)])lied to Bone Marrow.

By Ci-.vf.s 3!t,-.vK Pi.u.m 2f.O

The Influence of Different Tem})ernlnres on the Action of Drugs

on Autonomic Effector Cell.s. By HAka.v Bydi.v 270

<tn the .Syntlicsi.s of Proteins in Rat by Dialy.sed Casein Dige.st.s.

By K. A. .1. Wkktu.ni) 279

Effect of Acetylcholine and .Adeno.sine Triplio.s])hate on Dcnervnted
Mii.-^ele. By Fitirz Bt'CiiTiiAi, and Geouc Kaul.so.v 281

Ensc. i, (20. VI. lO lO.)

A Modified Preparation of the Univer.^al Buffer De.scribed by
Teoreli. and .STE.viiAtiK.v. By SvE.v O.STi,i.v(i and Pekka Yik-

TA.'IA 289

A Note on the Biogenesi.s of Choline nnd Creatine. By Gu.vxar

Stei:.v.«hoi.t 291

The Splanchnic Efferent Outflow of Impul.se.s in the. Light of Ergo-
t-aniine Action. By Bo Gernaxht and Yncve ZorrEUitAN .... 301
On the Effect of Some Pigments and Rcdo.x Sy.stems on the Ke.s-

jiiration of .Animal Tis.sue. By Gcn.vap. .Stee.wshoi.t 318

Fttrtlier Investigation.s on the Effect of Adeno.sinc Trij)ho.sj)hnte
and Related Pho.'jihorus Compound.s on Lsolatcd .Striated
.Mu.*:cle Fibre.'. By Fp.m. BircHTiiAi., Adam Deutsck nnd F. G.

K:.aipei,s 32.")

R.ite of Renewal of Ribo- and De.so.xtTibo Nucleic .Acids. By

E. llAJi.MAP.STEK and G. IfEVE.sV 335

On the Purification of the Thinmin-Innctivnt ing Fi.sh Factor II.

By Gf.v.VAp. .Ir.uE.v 3M

The Gri'tric Lip.T-e in Man. By Fnm Schoniievdjip. and Kiiestek

Voi.QVAtnv, ’ 319

The Effect of Piijeridine and Allied .Sub.stance.s on Mammalian
.Skeh tal Alu.'eh;. By RtcHAiti) F. Oii.vell 301

VOIi. 11. INDEX.

I’ng.

Effect of Minute Amounts of Barium on Cardiac JIuBcle. By

ADAjr Deutscii and Gunnar Ld.vdi.v 373

The Principle of Evacuation of the Stomach in Infants and Pre-
matures. By Stephan Vendee 3S0

Supplcmcntum XXXIII. Contrihiitions to (he Knowledso of Exo-
genous Insulin on the Glycogen Storage of Normal Animals. By
Are Swens.son.

Supplcmcntum XXXIV. On the Pre.scnce, of Histamine in Plasma
in a Pln'siologically Active Form. By X'ii.,s E.m.'IEM.v,
Suppiementum XXXV. A Study of the Ke.spiratory Beflexes Elicited
from the Aortic and Carotid Bodies. By Bo IC, Ger.vandt.

INDEX AUCTORUM.

Pag.

Asmussen, E., Aerobic Recovery after Anaerobiosis 197

Astrup, T. and A. Fischer, Protein Metabolism of Tissue Cells 187
Astrup, T. and J. Piper, Fibrinogen and Polysaccharide Poly-

sulfuric Acids 211

Augustinsson, K.-B., Choline Esterases in Marine Invertebrates 141
Buchthal, F., a. Deutsck and F. G. Knappeis, Adenosine

Triphosphate on Isolated Muscle 325

Buchthal, F. and G. Kahlson, Acetylcholine and Adenosine

Triphosphate on Denervated Muscle 284

Deutsch, a., F. Buchthal and F. G. Knappeis, Adenosine

Triphosphate on Isolated Muscle 325

Deutsch, A. and G. Lundin, Barium on Cardiac Muscle 373

Euler, U. S. v., Sympatbin E Properties in Spleen Extracts . . 168
Fischer, A. and T. Astrup, Protein Metabolism of Tissue Cells 187

Gad, I., Phosphatase Activity in Serum and Organs 151

Gernandt, B., G. Liljestrand and Y. Zotterman, Efferent

Impulses in Splanchnic Nerve 230

Gernandt, B. and Y. Zotterman, Injury Discharge of Afferent

Nerve Fibres 248

Gernandt, B. and Y. Zotterman, Splanchnic Efferent Outflow

and Ergotamine Action 301

Hajimarsten, E. and G. Hevesy, Renewal of Ribo- and Desoxyribo

Nucleic Acids 335

Hartiala, K. and M. ELarvonen, Anoxia on HCl-Secretion .... 85

Hevesy, G. and E. Hamjiarsten, Renewal of Ribo and Desoxy-
ribo Nucleic Acids 335

Hofpmeyer, J., 0. Jalling and F. Schdnheyder, Serum Phos-
phatase Activity 160

Jalavisto, E. Perception of Weight Ill

Jalling, 0., J. Hofpmeyer and F. Schdnheyder, Serum Phos-
phatase Activity 160

JuuRUP, A. and L. Muido, Effects of Cigarette Smoking 48

Kahlson, G. and F. Buchthal, Acetylcholine and Adenosine

Triphosphate on Denervated Muscle 284

ItjVRvoNEN, M. and K. Hartiala, Anoxia on HCl-Secretion 85

Knappeis, F. G., F. Buchthal and A. Deutsch, Adenosine
Triphosphate on Isolated Muscle 325

INDEX AUCTOBUJI.

Kkdhoffer, P., Inulin in Urine and Plasma

Kbuhoiter, P., Inulin as Indicator for Extracellular Space . . .

Kruhoffer, P., Solutes in Interstitial Space • • • • •

Lidjestrand, 6., B. Gernandt and Y. Zotterman, Efferent

Impulses in Splanchnic Nerve

Ldndin, G., g-Strophantin on Cardiac Muscle

Lundin, 6. and A. Deutsch, Barium on Cardiac Muscle

Muido, L. and A. Juuruf, Effects of Cigarette Smoking

Munk Plum, C., Tissue Culture _

Piper, J. and T. Astrup, Fibrinogen and Polysaccharide Pol 3 'siil-

furic Acids

Rydin, H., Temperature on Action of Drugs

Schmidt-Nielsen, B., Microdeterminatiou of pll in Saliva

ScnJfiDT-NiELSEN, B., pH in Saliva

SenONHEYDER, F., J. IIoFF.MEYER and 0. Jalling, Soniin Phos-
phatase Activit}'

ScHONHEYDER, F. and K. VoLQVARTZ, Gastric Lij)ase in Man., i

Steensholt, G., Creatine in Animal Body

Steensholt, G., I\Icth}'lation Processes in Wheat Germs

Steensholt, G., Biogenesis of Choline and Creatine !

Steensholt, G., Ecs])iration of Animal Tissue !

UssiKG, H. H., Amino Acids in Haemol^unph

Vended, S., Evacuation of Stomach in Infants ;

ViRTAMA, P. and S. Ostli.vg, Modified Preparation of Universal

Buffer «

VoLQVARTZ, K. and F. Scho.vheyder, Gastric Lipase in Man . . 1

Wretlind, K. A. J., Synthesis of Proteins i

Zotterman, Y., B. Gernandt and G. Liljestrand, Efferent

Impulses in Splanchnic Nerve 5

Zotterman, Y. and B. Gernandt, Injury Discluirgc of Afferent I
Zotterman, \. and B. Gernandt, Splanchnic Efferent Outflow

and Ergotamine Action j

Agree, G., Thiamin-Inactivating Fish-Factor i

OiiNELL, R. F., Piperidine on Mammalian Skcletai Muscle t

Modified Preparation of Universal

Buffer

From the Institute of

Medical Physiology, University of Copenhagen.

Deteriniiiation of Imilin in Urine and Plasma.

P. KRUH0PFER.

Received 10 Octoher l94o.

The basis of all recorded determinations of inulin in sucli con-
centrations as occur in clearance test is a hydrol^is of the mate-
rial by acid reaction witli subsequent determination of the hydro-
lysis products. These products, which are mainly fructose, are
determined either by their reducing powers or colorimctrically
with certain colour reactions which arc mote or less specific for
fructose.

The colorimetric reactions described are a good deal more sen-
sitive than determination by means of the reducing powem of the
hydrolysis products; this fact of course suggests the use of the
colorimetric methods when the concentration in the fluids to ho
analj^ed is low or the available quantity of fluid is limited.

Another circumstance which equally reduces the usefulness of
methods based upon a reduction determination is that’ one often
finds rather considerable blank reductions in certain biological
materials (especially plasma), even after removal of glucose by
fermentation, other substances with a not inconsiderable power of
reduction being formed by the hydrolysis (e. g. hy glycogenolysisl).
For example, Van Slyke, Hiller and Miller (1935) in dog plas-
ma (precipitated with CdSOi/NaOH) found reductions equivalent
to 3—5 mg % inulin. Other authors have found 3—26 mg % in
man. If these values are correct, the blank reduction will form
a large part of tbe total reduction, unless very liigh inulin con-
centrations are used — which is undesirable, for one reason on
account of the expense. Consequently, variations in the value of

1 — 'i60215. Acta phys. Scandinav. Vol.ll.

P. KllDHOFFEIl.

fhe blank reduction •within the experimental period or errors in
its determination may involve considerable error in the plasma-
inulin determinations, even at concentrations of 40 — 60 mg %.

For these reasons, in the search for a suitable method I have
ignored procedures based upon reduction determinations and
have tested two of the colorimetric methods described in the
literature.

A brief description of the colorimetric methods employed for
inulin analyses may be given. They are based upon one of the
following three foundations;

1. Diphenylamine reaction: Heating of fructose ■with diphe-
nylamine in a solution containing hydrochloric acid leads
to the formation of a blue colour.

The most -widely used methods employing this principle are
described by Cobcoean and Page (1939) and Alving, R-dbin
and JiliLLER (1939). A feature common to the procedure of these
two groups of authors, in contrast to certain earlier methods using
the diphenylamine reaction is the following: Hydrolysis and colour
development proceed simultaneously. Ethylalcohol is added be-
fore hydrolysis and colour development, whereby the colour
compound formed is field in suspension from the very moment
it is formed. Apart from differences in the protein precipitation,
the methods differ in the main only in the fact that the former
authors apply heat from a boiling water-bath to open tubes for
only 15 minutes, whereas the latter apply it for 60 minutes and
to closed tubes. The methods differ also in respect of the tech-
nique employed for fermenting the glucose (glucose gives the
same sort of colouration as fructose, though much less intense).
The last-named authors remove the interfering glucose by fer-
mentation in the plasma itself, a yeast-cell suspension being ad-
ded; the resulting plasma dilution is corrected by means of a hae-
matocrit determination on the yeast-cell suspension. Cobcoban
and Page on the other hand ferment the glucose in the filtrate
from the protein precipitate (Somog-xi’s ZnSOi/NaOH). Spuh-
LER (1943) and Jensen (1942) retain^ Alving, Rubin and hliL-
ler’s method almost unchanged, though Jensen prefers pro-
pylalcohol to ethylalcohol.

2. Sebwanoff’s reaction:, Heating of fructose -with resorcin
in a solution containing hydrochloric acid leads to the
formation of a red colour.

ISUITH IK VRIKE AKE PJiASMA.

A modification of tliis method has been employed by Bob
( 1934:) for determining fructose, and Steikitz (1938) utilized his
modification for inulin determination. Another modification of
Sewwanoff’s reaction for inulin determination is described by
Hatz and Szecsenyinagi (1940). In both modifications hydro-
lysis and colour development proceed simultaneously, but in
Steinitz’s method this takes place by heating to 80° C. for 8
minutes, vhereas the other authors use 10 minutes, heating in
a boiling water-bath and — in contrast to Steinitz only then
adding alcohol.

3. Vanillin reaction; Heating of fructose with vanillin in a
solution containing hydrochloric and phosphoric acid gives
a read colour. Habeay (1942).

In the search for a suitable method the following plan of work
was observed; First of all, by emplojdng pure aqueous inulin solu-
tions an endeavour was made to find the most exact colorimetric
method. Ha^ung found a satisfactory method, tliis was employed
for the elucidation of certain other problems, viz.;

1. Becoverj’’ percentage of inulin in the filtrate after various
precipitations. 2. Eccovery percentage of inulin after fermentation.
3. The blank value in plasma and urine.

Colorimetry was performed with a Weka photoelectric colori-
meter (a two-cell apparatus on the substitution principle des-
cribed by Havemanxt (1940)) using an incandescent bulb and
colour filter.

Colour Dovclopmcnt on Aqueous Inulin Solutions.

Various modifications of the diphcnylaminc and Soliwanoff
reactions have been tested.

a) Diplienylaminc reaction: Both Cobcokan and Page and
Aevtng, Eubin and Milleb employ the following procedure;
To a solution containing 80 ml. cone. HCl and 110 ml. abs. ethyl-
alchol, 10 ml. of a 10 % solution of diphcnylaminc in abs. ethyl-
alcohol is added immediately before use. Mix two parts of this
wth one part of inulin solution. As it was found, however, that
the blank tests gave an increasing seagroen tint when one used

mixed for some time, or ah
old diphenylamine solution, the reagent was prepared fresh every
day from its three constituents. ^

As the colour compound formed on heating 4—10 mg % inulin

P. KRUH0FFER.

solutions mtli tins reagent displayed a tendency to precipitate
as minute particles, tlie alcoliol concentration was raised to 65 %
of a 96 % ethylalcoliol witli corresponding reduction of tlie hy-
drooliloric acid content. However, . even witli a reagent of tliis
composition (500 mg. diplienylamine dissolved in 65 ml. 96 %
ethylalcohol, with 35 ml. cone. HCl then added) a slight preci-
pitation was observed now and then.

In advance Corcoran and Page’s procedure, with the heating
of the mixture in open tubes, was considered to be unsuitable
on account of the possibility of differences in the evaporation
of alcohol and HCl in the various tubes, and in fact tentative tests
confirmed that supposition. In experiments with Alving et al’s
type of tube, with a screw lid and rubber liner, there was a slight
turbidity in the blank samples, although the rubber liners were
thoroughly cleansed in soda lye; it was therefore decided to heat
the mixture in sealed and carefully cleansed test-tubes in all
experiments. The tubes were heated in a boiling water-bath.

Plotting the absorptions (measured with red filter RGi) ob-
served, when an inulin solution is heated with twice the quantity
of the diphenylamine reagent in a boiling water-bath, as func-
tions of the boiling time, it was found that the colour absorption
increases quickly at first,’ then more and more slowly; but even
after 90 minutes’ “boiling” the colour development has not reached
its end-point. Longer heating was not thought practicable, and
therefore it was decided to stick to the 60 minutes recommended
by Alving et al. However, the circumstance that colour develop-
ment is still in progress has the effect that even slight temperature
differences in the waterbath (e. g. owing to different barometric
pressure on different days) cause the colouration to be somewhat
different.

On carr 5 ’ing out decuplicate analysis with the technique de-
scribed, all the samples being heated in the same water-bath,
such deviations were observed, that the method, though it may
serve for some purposes, cannot be considered ideal. The results
of a decuplicate analysis made under these conditions may be
given as an example: 4.89; 4.91; 4.96; 5.00; 5.01; 5.03; 5.04; 5.09;
5.16, Here the mean value is 5.006, the standard deviation 0.081
or 1.6 % of the mean value, and the standard error 0.025. Not
infrequently, however, even among analyses from the same water-
bath there are greater deviations than the above, especially in
a positive direction, so it is doubtful if the values show a normal

INULIK IN 0EINE AND PLASMA. «>

distribution. The cause of this is not quite known; but as the
diphenylamine reagent gives deep blue to ^een tints with a
number of oxidizing agents (chromates, potassium permanganate,
nitrates, etc.), the most scrupulous cleanliness with all glassware
is necessary. Tor analyses heated in different water-baths and
especially at different times there is usually greater deviation.
Experiments with the use of propylalcohol instead of cthylalcohol
made no improvement.

b) Seliwanofj’s reaction: lATien testing this reaction the proce-
dure described by Roe (1934) was followed. Various proportions
of the ingredients in the resorcin-ethylalcohol-hydrochlorid acid
reagent were tried until the following was decided upon: 100 mg.
resorcin dissolved in 60 ml 96 % cthylalcohol, whereafter 40 ml.
cone, hydrochloric acid was added. One part of inulin solution
(2.5 ml.) -j- two parts (5 ml.) reagent were heated in scaled test-
tubes.

A series of tests of the same concentration Averc heated in a
water-bath at 100° respectively 80° C. for various periods and
then, after quick cooling to room temperature, the absorptions
were measured with green, blue-green and blue filters.

Generally speaking, there is first a successive development of
a cherry-red colour, which gradually tones into a lemon-yellow
tint, which remains constant if the heating is continued. At 100°
this colour development proceeds at a much greater velocity than
at 80°, and the process is quickest at high concentrations of
hydrochloric acid. In correspondence Avith tliis colour develop-
ment it is observed that the colour absorption, measured with
the aforesaid filters, first rises to a maximum and then gradually
decreases to a constant minimum. The figure 1 illustrates these
changes when the mixture is heated in a boiling AA'ater-bath and
colorimetry undertaken immediately after cooling. Under such
conditions the maximum absorption occurs after about 10 minutes;
at 80° only after 15 — 20 minutes.

With the methods described in the literature, colorimetry is
undertaken at times when there is still a red tint, but a test shoAved
that this red tint, appearing for instance after 10 minutes at 100°
is not constant, as the solution after cooling shows a gradual
increase in the colour absorption. Even Avhen colorimetry is under-
taken immediately after cooling, the values obtained from a
decupheate analysis Avith 10 minutes’ boiling show greater deviation
than in the above example of the diphenylamine reaction.

P. KmJH0FFEIl.

The various methods in the literature, rvith colorimetry on red,
having failed to give satisfactory results, the possibility of ap-
plying colorimetry at the subsequent minimum, corresponding
to the lemon-yellow tint was examined. To save time the tests
were made only after heating in a boiling water-bath, because
at lower temperatures the process takes longer to reach the end-
point. In all the following readings a blue filter was used with
a maximum transmission at about 4,500 A and a cuvette with
a layer thickness of 5 mm. The figure 1 shows, that at 100° the
minimum is reached after about 45 minutes, but a period of
60 minutes was decided upon in order to be on the safe side.

units

The following results show the accuracy obtainable by this pro-
cedure: 20 values obtained, 10 on each of two days, the analysis
being made on the same inulin solution: First day: 4.94; 4.97;
4.97; 4.98; 4.99; 4.99; 5.00; 5.00; 5.02; 5.06, ^vith the mean value
at 4.992. Second day: 4.96; 4.98; 4.98; 4.99; 5.00; 5.00; 5.01; 5.02;
5.06; 5.08, with the mean value at 5.008. The standard deviation
for the two series together is 0.035, or 0.7 % of the mean value;
the standard error is 0.008. This shows that the method permits
of determining inulin in aqueous solutions with satisfactory
accmacy. As colour development has reached its end-point after
60 minutes’ heating, the heating time is not critical as in the
diphenylamine method, nor has it any effect on the resiilts that
the temperature of the water-bath is not exactly 100° C.

Like the diphenylamine reagent, the resorcin reagent is extre-

INULl!^ IM tJUlKE AKD PLASMA.

mely sensitive to oxidizing agents, for wliich reason the utmost
cleanliness vdth all glass is necessary in this case also. It is also
■sensitive to ultraviolet rays, and therefore it must on no account
he exposed to bright sunlight.

It was observed earlier (Corcokax and Pack; Alving, Burix
and SfiLLER) that glucose gives the same sort of colour develop-
ment, though less intense than fructose (inulin) with diphenyl-
amine reagents. Much the same occurs in the resorcin reaction
described above. The following table (1) shows that glucose gives
a colour development corresponding to 7.9 % of the colour devel-
oped by the same amount of inulin. This of course means that
when employing the resorcin method with biological fluids con-
taining glucose wc must first remove the glucose or correct for
the colour development it entails.

Table 1,

Glucoso

concentration

mg %

Equivnient to
an inulin cone,
nig %

Thus 1 mg. inulin
is equivalent to
mg. glncoso

8.125

O.GC

12.3

1G.20

1.27

12.fi

24.375

1.92

12.7

32.5

2.58

12, c

C5.0

5.11

12.7

Conclusion: On nn nverngo 1 mg. inulin i.s cqiiivnlcnt to 12.(55 mg. glucose.

Becovery*X>' of Inulin in the Filtrate from Tarions Protein
Precipitations on Plasma ami Aqueous Inulin Solutions,
and the Rccovery-fo of Inulin after Fcrmcntivtion.

Regarding the suitability of various methods of protein preci-
pitation the literature contains a number of contradictory opinions.
The methods employed in inulin determinations have e.spccially
been Zn{OII)s-precipitation a. m. So.mogyi (1930, procedure 2)
or Cd(OH)j-ptecipitation a. m. Fujita and Iwatake (1931).
Opinions are divided on the subject of the recovery of inulin by
these precipitations. For example, Ar,vixo, Ruj 5 in and Mh-ler
( 1939) consider that both methods give “identical and correct
values”. Steixitz (1938) finds that the Cd(OH), -precipitation
resets m a lower recovery than the Zn(OH), method, and appa-
rently he considers that the latter gives a quantitative recovery
m he filtrate. Similar views were advanced by Eappaeort (1937).

For the protein precipitation the present author emplo 3 md

P. ERUH0FFER.

SoMOGYi's (1930, Procedure 2) Zn(OH )2 metliod, althougli Laving
regard to tLe yeast blank value (v. i.) the reagents were adjusted
to eacL other a little differently (60 ml. ZnSOj/HjSOj reagent was
titrated with pLenolpLtLalein as indicator with 6.4— 6.5 ml.
NaOH, about 0.79 n).

Por precipitation pipette 8 ml. ZnS 04 /H 2 S 04 reagent, add to
this 1 ml. analysis fluid (plasma, urine etc.) and finally 1 ml.
NaOH; mix tLorougMy, centrifuge.

Por fermentation I preferred tbe procedure of Corcoran and
Page (1939), with fermentation in tbe filtrate from tbe Somogyi
precipitation, to fermentation in tbe plasma itself. Tbe reason
was that with tbe small quantities of plasma usually available
(1 — ^2 ml.) we must expect a not inconsiderable dilution error by
direct fermentation on plasma by Axving, Burin and IMiller’s
method.

In preliminary tests on solutions of glucose in tbe filtrate from
Somogyi precipitations it was found that 2 ml. of a 10 % washed
yeast suspension (Baker’s yeast from Tbe Danish Distilleries)
sufficed to remove at least 150 mg % glucose from 7 — 8 ml.
Somogyi filtrate in 30 minutes at 30 — 35° C. Permentation was
therefore carried out in tbe following manner: 2 ml. 10' % freshly
washed yeast suspension was pipetted into centrifuge tubes and
centrifuged for 20 minutes at 3,000 r. p. m., whereafter tbe super-
natant fluid was removed completely by suction; on to tbe remain-
ing yeast was poured tbe supernatant fluid from tbe Somogyi
precipitation, and tbe yeast stirred until it formed a homogeneous
suspension; after standing for 30 minutes at 30 — 35° tbe yeast
was again centrifuged off and tbe clear supernatant fliud was used
for tbe colour-reaction as described above.

Por tbe recovery test we employed aqueous solutions and
plasma with a known inulin content from which tbe glucose
bad been fermented.

Tbe procedure was as follows: S being an aqueous solution of
inubn of known concentration, tbe following solutions were pre-
pared. 1) Water. 2) 1 ml. S -f- 10 ml. water. 3) 1 ml. water -f- 10 ml.
fermented plasma. 4) 1 ml. S -f 10 ml. fermented plasma. Six
samples of each were precipitated as above a. m. Somogyi; three
were taken from each group for fermenting in tbe manner de-
scribed, whereas tbe other three were not fermented. In addition,
5) 1 ml. water -1- 9 ml. Somogyi filtrate, and 6) 1 ml. of tbe mixture
2 9 ml. Somogyi filtrate.

INUm IN URINE AND PLASMA.

Tftblo 2.

P. KRDH0FPER.

Bl.auk Yalue in Plasma and Urine.

When employing Zn(OH )2 precipitation as described by SoMO-
GYi (50 ml. ZnS 04 /H 2 S 04 reagent titrated with 6.7 — 6.8 ml. NaOH
and then carrying out the fermentation as described above, it
was found that Somogyi filtrates of both fermented plasma and
water gave blank values of a not inconsiderable order. It was
observed, however, that by using a somewhat stronger NaOH
for the precipitation the blank value could be greatly reduced.
This must be due to the yeast gi^ng off less chromogenic material
at the more alkaline reaction thus conferred on the Somogyi
filtrates. Table 3 illustrates this.

No. of ml. NaOH
used for titrating
50 ml. ZnS 0 ,/H.S 04

6.45

6.55

6.90

7.10

7.30

Table 3.

Blank value equivalent to mg % inulin
when precipitating when precipitating

water

0.2

1.2

3.9

7.8

8.2

fermented plasma
0.5
0.5
0.5
1.0
6,0

It appears from Table 3 that in order to get the blank value
pressed down to a minimum the NaOH solution used should be
so strong that only 6.4 — 6.5 ml. of it is used for titration. Numer-
ous analyses have shown that by this means one gets very low
blank values both for plasma (from rabbit and man) and for
urine (same objects) diluted to a degree suitable for clearance
tests. As a general rule values of less than 1 mg % chromogen
expressed as inulin are found for plasma, and as this value varies
very little in the course of several hours, the blank value repre-
sents no perceptible source of error with plasma-inulin concentra-
tions of about 60 mg %. As to the urines, if these are diluted to
such a degree, that the inulin concentrations will be approxi-
mately the same as in plasma (for a healthy adult man to about
120 ml for each minute the urine is collected), one usually finds
still lower blank values in a Somogyi filtrate.

The Method in Practice.

Two elaborations of the method are given below. The first one,
in which a fermentation is avoided, a correction for the chro-
mogen effect of the glucose being inserted instead, is best for

INULIN in UniNE AND PDASMA.

glucose coucentrations beloiv 150 rag % and undoubtedly ^vcs
the more exact results when tlie glucose concentrations are low.

1) Procedure without fermentation.

Reagents-. Resorcin. Dissolve 100 i 1 rag. purest crystalline
resorcin in 60 ml. 96 % etliylalcohol, then add 40 ml. cone. HCl
pro analysi. Shake and cool. Prepare fresh for each series of

analyses.*

Somogyi precipitation fluids;

a) Add 12.5 g. ZnSO^ pro analpi to 31.25 ml. n/1 HeSO^ per
analysis and make up to 1,000 ml. with distilled water.

b) About 0.79 n NaOH.

MTien titrating with phenolphthalcin as indicator 50 ml. of
a) should take 6.4— 6.5 ml. of b). (Of course, in this procedure
one may also use the original strength of NaOH proposed by
Somogyi.)

Centrifuge tubes prepared with heparin: To every 10 ml. centri-
fuge tube add 0.1 ml. 500 mg % heparin (Leo) = 125 Howell
units; dry — turning occasionally — in a desiccator at 60 — 80°
(or over a radiator). This quantity of heparin will definitely pre-
vent the coagulation of 10 ml. blood for five hours at room
temperature.

Reagents (except precipitation reagents) for blood-sugar tests
a. m. Hagedoun and Norman Jensen (1923).

Procedure: Into a 10 ml. conical centrifuge tube pipette 8.00 ml.
precipitation fluid a). With a pipette calibrated to contain 1.00 ml.
add plasma (or diluted urine), sucking up and blowing out repeat-
edly. Finally, add 1.00 ml. precipitation reagent b).

Mx the contents thoroughly by turning on a clean finger;
stopper and allow it to stand at least 10 minutes — or eventually
over night.

Centrifuge twice five minutes, the interval between the two pro-
cesses being employed for shaldng gently in order to wash dowm the
sediments adhering to the surface and the upper walls of the tubes.

Of the clear supernatant fluid pipette samples for determining
glucose and total chromogen:

1) For glucose determination pipette a suitable quantity of
ilmd (usually 1 ml.); add this to 10 ml. distilled water in a Ha^e-
dorn boiling tube; then add 2 ml. of Hagedorn's KjFeCNe reagent.
whereafter the glucose analysis proceeds in the usual manner.

a glat'stoweS bSf ^bia reagent shonld be prepared in

glass stoppered bottle. The three ingredients shonld be iept in the same way.

P. KRUH0FFEK.

If 1 ml. of the fluid -was employed the ordinary Hagedorn-Jensen
table shows directly the content of glucose in plasma (or diluted
urine) in mg %.

2) For determining the total chromogen (inulin + glucose)
the procedure is as follows: Of the supernatant fluid pipette
2.50 ml. into a perfectly clean test tube 15 X 150 mm. (cleansed by
boiling in about 5 % hydrochloric acid-alcohol, followed by thor-
ough rinsing in distilled water; the hydrochloric acid-alcohol can
be used several times). Then with a Krogh’s syringe-pipette add
5.00 ml. resorcin reagent. In doing so one should carefully avoid
to place any reagent so high up in the tube that it becomes heated
during the following sealing process. The tube is then sealed in a
blow-flame. The tubes are heated for 60 minutes in a boiling water*
bath, then cooled off to room temperature in running water.
Then open the tubes (with a glass-file first, then break off). Before
being opened they may stand at least an hour in a dark place
without changing colour.

Colorimetry is carried out as already described in a photoelec-
tric colorimeter with a blue filter in cuvettes with a layer thick-
ness of 5 mm. For zero setting a sample with 2.50 ml. distilled
water -j- 5.00 ml. resorcin reagent boiled for 60 minutes is used.
Then take another sample which has passed through the entire
analysis procedure as described, but in which 1 ml. distilled
water is used instead of plasma (urine); this gives the blank value
of the precipitation reagents.

How much the colour developed corresponds to in mg % inulin
is read from an adjustment curve. This is constructed on the basis
of the colours developed by pure aqueous inulin solutions when
2.50 ml. are boiled for 60 minutes in 5.00 ml. resorcin reagent.
(It is convenient to employ the folloiving concentrations: 10;
7.5; 6.25; 5.0; 3.75; 2.5; 1.25; and 0.612 mg %.) The concentra-
tions are plotted as the ordinate, the number of units on the revolv-
ing drum of the colorimeter as the abscissa. As the mixture is
diluted 10 times during precipitation, the chromogen concentration
in plasma (or diluted urine) is ten times as high as that read from
the curve.

The inulin concentration is calculated as follows:
a) in plasma:

mg % inulin = measured chromogen concentration (in mg %

■ T ^ 100 , . .

mulin) glucose concentration in mg % ‘0.079.

INDTitN IN UKINE AND PLASMA.

b) in diluted urine:

mg % inulin = measured cliromogen concentration (m mg /o

inulin)’^^ — glucose concentration in mg %• 0.079.

The fact that the above procedure gives reliable results will
be seen from an example such as the following: To 10 ml. fer-
mented plasma 1 ml. of a solution containing GCO mg % inulin
and 1,100 mg % glucose was added. By means of the above
analysis procedure and calculation the following inulin concentra-
tions have been found in the resulting mixture: 59.8; 60,0; 60.2;
60.4; 60.5 and 60.8.

2) Procedure with jermentation. This can be used with plasma-
glucose concentrations up to 2,000 mg %.

Reagents. These are the same as those employed in Procedure 1.

Further a 10 % yeast-cell suspension: 50 g. fresh baker’s yeast
(Danish Distilleries) is washed three times with distilled water
and then filled up to 500 ml with dist. water. "Will kecii fourteen
days in a refrigerator.

Preparing jermentation tubes. These must be prepared fresh for
every series of analyses. Of the above yeast suspension take a
suitable quantity, which must again be washed three times before
use. Pipette 2 ml. of the freshly washed suspension into 10 ml.
conical centrifuge tubes with rounded bottoms. Centrifuge the
tubes for 20 minutes at 3,000 r. p. m. Immediately after centri-
fuging draw off the supernatant fluid completely with a capillary
tube in conjunction with a water suction pum]>.

Procedure. Plasma (or diluted urine) is preci])itat cd in the same
manner as in Procedure 1. Centrifuge for 10 minutes at 3,000
r. p. m.

The whole of the clear liquor above the sediment is poured into
one of the fermentation tubes which have just been prepared.
(If insignificant, quantities of sediment arc carried over it is of
no consequence.)

With a small glass rod stir the yeast in the fluid until it forms
a homogeneous suspension, then stopper the tube and allow it
to stand — ivith occasional gentle shaking — for 30 minutes in
a waterbath at 30-35°. Then centrifuge the stoppered tube for
10 minutes at 3,000 r. p, m.

supernatant dear fluid now pipette 2.50 ml. into a clean

Sal “t’oCr

P. KRUD0FFER.

The blank values for fermented plasma or diluted urine are
determined in the following manner: 1) for Plasma: Plasma
taken prior to the inulin injection is treated according to the
method described above. 2) Diluted urine from a pre-period is
treated in the same manner. (Carry out the dilution so that urine
collected over a period just as long as the experimental periods
is diluted up to the same volume as the urine from those periods.)
The zero adjustment is the same as in Procedure 1.

The calculation of the inulin concentration proceeds as follows:

in plasma:

Inuhn concentration = (measured inulin cone, in mg

plasma blank value in mg % inulin)

100

95 ^'

% -

b) in diluted urine:

Inulin concentration = (measured inulin cone, in mg % —
blank value in diluted urine in mg % inuhn) .

•74

This procedure also gives fairly accurate results; the following
may serve as a representative example: To 20 ml. fermented
rabbit plasma 2 ml. of an aqueous solution with a content of
468 mg % inuhn and 5 % glucose were added. With Procedure
2 the following values for the inuhn content were found: 41.5;
41.7; 41.9; 42.0; 42.1; 42.1; 42.2; 42.4; 42.5; 42.6. This gives a
mean value of 42.1 against a real value of 42.5, or a difference of
about 1 %. The values group about the mean value with a stan-
dard deviation of 0.37 or about 0.9 % of the mean value.

Summary.

A colorimetric method for the determination of inuhn in plasma
and urine is presented.

It rests on the measurement of the yellow colour (absorption
measurement at 4,500 A) developed by heating inulin at 100°
for 60 min. with resorcin in a solution containing ethyl alcohol
and hydrochloric acid. This marks a real end point in the colour
development process; the resulting colour compound is very stable
and, contrary to the blue diphenylamine colour, it shows no
disposition to precipitate.

INUWN IN URINE AND PLASMA.

The recovery percentages of inulin after protein precipitation
and fermentation procedures have been determined. Further
the blank values of plasma and urine.

Two procedures — ■ one with and another without ferment-
ation — are elaborated.

Eeferenees.

Alving, a. S., J. Rubin, and B. F. Miller, J. biol. Chem. 1939. 127.
609. . .

CoBCOKAN, A. C., and J. H. Page, Ibidem 1939. 127. 601.

ProiTA, A., and D. Iwatake, Biochem. Z. 1931. 212. 43.

Hagedorn, H. C., and B. N. Jensen, Ibidem 1923. 135. 46.

Harlay, V., J. Pharmacie 1942. 9. S. 2. 251.

Hatz, E. B., and L. Szecsenyi-Nagi, Biochem. Z. 1940. 30G. 71.
Haveuann, R., Ibidem 1940. 306. 224.

Jensen, E., Bibl. Lieger 1942. 134. 175.

Rappaport, E., Klin. Wschr. 1937. 16. 1358.

Roe, j. H., J. biol. Chem. 1935. 107. 15.

Slyke, D. D. V., A. HrLLER, and B. F. Miller, Amer. J. Physiol.
1935. 113. ,611.

SoMOGYi, M., J. biol. Chem. 1930. 36. 655.

SpOhler, 0., Dtsoh. Arch. Min. Med. 1943. 190. 20.

Sxeinitz, K., j. biol. Chem. 1938. 126. 589.

From the Institute of Medical Physiology, University of Copenhagen.

Iniilin as an Indicator for tlie Extra-
cellnlar Space.

PODL KRUH0PFER.

Received 10 October 1945.

Various substances, some present in tbe organism, others to
be injected, have been proposed for tbe measurement of tbe
extracellular volume in tbe whole organism as well as in individual
tissues, tbe last named determinations often being carried out to
calculate tbe electrolyte composition of tbe intracellular fluid.

Tbe common basis on wbicb tbe calculations of tbe extracellu-
lar volume rests is tbe assumption that, when equilibrium bas
been attained, these materials will be present in tbe interstitial
fluid in just tbe same concentration as in plasma idtrafiltrate.
Different investigations support this assumption (Peters (1935)
and Maurer (1938)).

An ideal indicator of tbe extracellular fluid space should pos-
sess tbe following properties:

1) When sufficient time bas elapsed for complete distribution
tbe material must be found everywhere in tbe extracellular fluid
and exclusively there and tbe concentrations of it must be uni-
form throughout tbe interstitial fluid and throughout plasma
water, tbe ratio between tbe concentrations in tbe two compart-
ments being a known fixed figme.

2) It must not be ebminable from tbe extracellular space (by
excretion, storage, or decomposition).

Further if it is a foreign substance to be injected, tbe quantity
wbicb must be administered to get concentrations wbicb may
be determined with sufficient accuracy must

inulin as an indicator for the bxtkacellddar space. 17

3) Dot possess so Wgli an osmotic pressure, tliat it causes a
perceptible displacement of water from tte intracellular to the
extracellular space.

4) not be toxic. _ • j.

llTien complete distribution has been attained for an indicator

of this kind, the extracellular fluid volume may be calculated as
follows:

(1) For a non-electrolyte;

extracell. vol.=
(in ml.)

quantity injected (or present in the organism)
quantity in 1 ml plasma water

(2) For an electrolyte ion:

quantity of ion inj ected (or present in the organism)^
extrawU^ol quantity of ion in 1 ml plasma water • F

ion concentration in interstitial fluid

where F == 7~r< = 5 1 —

ion concentration in plasma water

If a substance is not known to be exactly confined to the
extracellular space the neutral term “available space” should be
preferred.

If we use these formulae for calculations before the distribution
■is complete, the figure calculated will vary with the point of
time for the determination. These values we may name “■volumes
of distribution”.

Naturally, to calculate a volume of distribution at a moment
when the distribution is not yet complete may be considered
illogical; nevertheless it has proved useful to do so, as a curve
in which volumes of distribution are plotted against time (curve
of distribution) represents a simple method of graphic illustration
of the rate of distribution, moreover such curves make a direct
comparison between the rate of distribution of different solutes
possible.

The volume of distribution for a non-electrolyte is defined as
the volume which should act as solvent for the injected quantity

_ ^ The precise expression here irould bo

Mj ected quantity — plasma volume • cone, in plasma

, quantity in 1 ml plasma water -E + volume,

however, the above given formula is a good approximation.

^—460213. Acta phys. Scandinav. Vol. 11.

POUI. KRUH0FFER.

if tlie concentration -was to be tbe same as in plasma water
tbrougbout tbe volume. Actually, before tbe distribution is com-
plete there must of course always be places in wbicb tbe con-
centration is less and eventually also places in wbicb it is bigber
than in tbe plasma water.

In practice, it bas proved impossible to find an indicator sa-
tisfying all tbe above mentioned requirements; at least every
proposed substance is excreted in tbe urine.

Certain authors have therefore adopted tbe procedure of deter-
mining tbe urinary excretion of tbe substance administered in a
single dose, and then in tbe calculation according to formula
(1) or (2), replacing tbe injected quantity with tbe quantity re-
tained at tbe end of tbe experimental period. In such a procedure,
however, on account of tbe continuous excretion, an equibbrium
between the concentrations of tbe indicator substance in plasma
and interstitial fluid is never attained. For this reason tbe plasma
concentration found is not a correct measure of the distribution;
for instance, on a falling plasma concentration when this bas
passed tbe average interstitial concentration, tbe extracellular
space is estimated too high. 'Wlule errors from this sorrrce are small
for substances (as SCN~), which are excreted slowly, they may
become considerable for substances (inulin, sucrose) wbicb are
excreted quickly.

Therefore, if one wants to determine the available spaces for
substances of this last categorj’’, a procedure must be used in
which an equilibrium between plasma and interstitial fluid can
be established. Theoretically two methods should be possible;

1) After a large initial dose tbe substance might be administered
by steady, continuous infusion to keep tbe plasma concentration
constant until distribution was complete.

2) Tbe organs of excretion (Tbe kidneys) might be removed
prior to administration.

In practice, for quickly excreted, slowly diffusing substances
the first possibility -will be found unsuitable: at tbe end of a period
long enough for complete distribution, it is found that tbe quantity
retained represents only a small fraction of tbe quantity adnain-
istered and therefore can be determined only with insufficient
accuracy. Another soTirce of error in tbe determination of tbe
quantity retained is due to tbe fairly large quantities of these
substances which, owing to their high urinary concentrations are
accumulated in tbe urinary tract.

ISraS .S AK ISMCAIOB ESIBACELLOLAB SPAOB.

Kow, in tlie course of some experiments on rabbits,
determinations of inubn clearance by continuous infusio ,
imuression was received that tbe inubn available volume was
si^ficantly smaller than tbe values given in tbe literature for

tb? fmther investigation of this question, only method
2 appeared to be suitable. Hence tbe distribution of inubn in
nephrectomized animals bas been investigated. As a basis fox
comparison it was decided to determine tbe distribution ol sulpbo-
cyanate, mainly because this is tbe substance most frequently
employed by earUer workers.

Experimental Technique.

The animals used in the experiments were 13 amytalnarcotized
(60—70 cg/kg.) male rabbits of 2.5 to 3 kg. and an ether-morphia
narcotized male dog of 12 kg; prior to the experimmt, except in one
special case, the animals had been in a state of inanition for 18 hours,
hut with unlimited access to water.

On the day of the experiment the narcotized animals were strapped
on their backs on an electrically heated operating-table. The left
jugular and the right carotis were exposed and furnished with cannulae.
Both kidneys were then removed through an abdominal section, where-
after the abdominal wall was catefuUy sutured.

About 5 — 6 ml. arterial blood was drawn for determinations of the
blank value for the inulin and thiocyanate analysis. Then a solution
qf inulin and NaSCN in physiological NaCl-solution was injected
intravenously in the course of about 15 seconds; as a rule about 300 mg.
inulin and about 200 mg. NaSCN were injected in experiments on rab-
bits. The experimental period was reckoned from the middle of the
injection time. At suitable intervals samples of 5—6 ml. arterial blood
were then taken, up to a total of 8. The blood was stabilized with
heparine. The samples were centrifuged immediately after being seonred
and the plasma was pipetted off and employed for determining the
concentrations of inubn and SON".

_ In three of the experiments a test was also made of the effect of
^usmg a rather large volume flOO-^120 ml) of a sulphate-substituted

chloride concentration was determined before
infosion and m the subsequent blood samples

after the termination of the

POUL KRUH0FFER.

Technique of Analysis.

In blood plasma: The method employed for determining mulin
■was that published recently by the author (Kruhoffer 1945 a Pro-
cedure 2).

The SCN^ concentration -was determined by a slight modification
of Crandall and Anderson’s method (1934): Pipette 7 ml. distilled
■water into a 10 ml. centrifuge tube; ■with a pipette calibrated to con-
tain add 1 ml. plasma under repeated suction and blowing, then add
2 ml. 30 % trichloracetic acid, mix thoroughly, leave for 10 minutes,
centrifuge. Of the clear supernatant fluid pipette 3 ml., to which im-
mediately before colorimetry add 4 ml. of Crandall’s reagent (50 g
ferrinitrate + 25 ml. cone. HNO3 -j- distilled water to 1 litre). Exactly
1^/2 minutes after adding the reagent take the colorimetric reading. (In
the experiments here described eolorimetry was carried out in a photo-
electric absorptometer — Weka — described by Havemann (1940),
using a green filter BG, and a cuvette layer thickness of 5 mm.) Bead
the result from an adjustment curve designed for known aqueous solu-
tions (adjusted to silver nitrate), known concentrations being plotted
along the ordinate and the corresponding readings along the abscissa.
For plasma multiply by 10 on account of the dilution by deproteiniza-
tion, and introduce a correction of — 1.5 % for the concentration of
the supernatant fluid due to the fact that the protein sediment has
a certain volume.

The chloride in plasma was determined a. m. Keys (1937).

For peritoneal fluid the procedure was centrifuging to remove the
small content of erythrocytes and then analyzing exactly as for
plasma.

The excised tendons were immediately freed of muscle, cut into
small pieces and, after freezing in liquid air, pounded in a mortar.
A suitable quantity was weighed off into a centrifuge tube and pre-
cipitated in the same manner as plasma. After standing for 18 hours
in the refrigerator the sediment was centrifuged off and the supernatant
fluid was analyzed for inulin and SCN“.

Determining the Ratio of the Diffusion Coefficients for SGN~ and
Inulin. The technique here was that described by Northrop and An-
son (1929). The apparatus was a diffusion chamber with a volume of
24.7 ml. pro^vided with a Jena glass filter disc No. G4 with an active
diameter of 30 mm. and a thickness of about 2 mm. The diffusion ex-
periments were made at 37° C. The inulin employed was the same as
that in the animal experiments, a triple alcohol-reprecipitated material
from Schering-Kahlbatjm: “Inulin reinst”. The inner fluid employed
was an aqueous solution containing 1,500 mg % inulin and 700 mg %
NaSCN.

Under these conditions the ratio between the diffusion coefficients
Ug0.^ —

— proved to average 5.00 (± 1 %).

“^inulin

INOLIN AS AN INDICATOR TOR THE

EXTRACEIiDUDAR SPACE. 21

Calculations.

The formula given previously (1) rvas employed for calculating tlm
distribution volumes for inulin, the concentration in plasma water was

calculated as plasma cone. •

The calculation of the available volume for SON" was made nccord-
ine to formula (2), the same correction being made for the water con-
tent in plasma, whereas it was elected to put F = 1 . This latter cboicc
was made despite the fact that experiments by Rosenbaum and J^A-
viETES (1939) and the author’s own tests of peritoneal fluid showed
that F < 1. By taking F = 1, however, the chance of estimating the
SCN“ available space too high was definitely avoided.

In the experiments where the chlorides of the organism were diluted
by sulphate infusion, a volume containing chlorides capable of being
diluted (named "volume of dilutable chlorides”) was calculated accord-
ing to the following formula;

Cl-vol.t, X plasma-Clo = (Gl-vol-t -I- R) X plasmn-Clt,

where plasma-Clo means the plasma-chloride concentration prior to
the infusion; plasma-Clt the same the time t after termination of the
infusion; Cl volt.t the volume of dilutable chlorides c.'ilculntcd from tlic
dilution of plasma chloride existing at the time t, and N the infused
number of ml. sulphate solution.

Results.

Fig. 1 shows the result of an experiment in which a rabbit
(No, 17) was intravenously injected with 19G mg. inulin and 193
mg. NaSCN in 10 ml. fluid. The falling curves represent the con-
centrations (in mg %) of inulin and K’nSCN = calculated for an
ultrafiltrate of plasma as functions of the time after the injec-
tion; the concentrations are read on the ordinate scale on tho
left. The ascending curves represent the volumes of distribution
for mulm and SON calculated in the manners described above-
they are read from the ordinate scale on the right, where their
values in per cent, of the body weight can also be found.

I evident that the volumes of distribution for SCN~ through-

out are larger than the simultaneous volumes of distribution for
not d‘ of time aro

6.6 times higher than for imib 7 -/ ^ conditibns is

signer than for inuhn. If we presuirio -(and it may bo

POUL KRUH0FPER.

Ultrafiltrate

done witli good approximation) that this value also applies to the
organism and to the concentrations employed in the animal
experiments, we can from the inulin distribution curve construct
the distribution curve according to which in theory SCN“ would
necessarily be distributed if finally it was distributed in the same
volume as inulin and both substances were distributed in this
volume by diffusion alone. The points on this theoretical distri-
bution curve for SCN~_ are found as follows: "When for instance
100 minutes have elapsed the inuhn has distributed itself over

100

420 ml., consequently at the time = 18 minutes SON'

0.6

should have distributed itself in the same volume.

On fig. 1 will be found the constructed, theoretical distribution

curve for SCN“ ( ). If the substances are distributed

not only by diffusion but also by convection (by movements of
fluid in the interstitial space), the theoretical curve of distribution
for SCN~ will lie. nearer the inulin curve. Thus it appears very
clearly that the thiocyanate ions distribute themselves in a much

ISUX.m AS AN INDICATOR FOR THE EXTRACEDLUDAR SPACE.

ultrafiltrate

larger than could be theoretically expected under the said condi-
tions. (In part of the distribution volume — i. e. in plasma — the
distribution does take place by convection; but as this proceeds
very rapidly, it will scarcely affect the course of the curves at all
after the first few minutes; of. EjRtJHOFpER (1945 b).)

Corresponding results were found in other experiments with
rabbits and in a single one with a dog. The results of the latter
appear from fig. 2. The dog, which weighed 11.8 kg was injected
intravenously with 1,200 mg. inulin and 685 mg. NaSCN in 20 ml.
fluid in the course of 10 seconds.

The data included in table 1 will give an impression of expe-
rimental results obtained in all the experiments performed. This
table shows the various distribution volumes as a percentage of
the body weight at the time t minutes after the injection of inulin
and NaSCN. The results at t = 30 and t = 150 minutes require

POUL KRUH0FFEU.

Table 1.

Disiribution imlumes in perceniage of the body weight.

Exp. No.

Animal’s

weight

(kg)

Vol. of distrib. for Inulin

Vol. of distrib. for SCN"

t = 0

t = 150

Rabbits

2.S

—

— >

—

27.5

2.S

—

26.5

3.0

—

14.5

—

2.9

—

14.5

—

2.5

14.5

11.5

15.5

24.5

2.8

—

2.5

28.5

2.9

13.5

26.5

22.5

3.0

—

16.5

—

3.0

16.5

25.5

2.0

—

2.5

16.5

29.5

31.5

Mean values

11.9

15.8

27.9

24.9

28.7

Rabbit

Dehydrat.

8.5

11.5

19.5

3.0

*—

Dog

9.5

—

14.5

—

22.5

27.5

no explanation. The results at t = 0 were arrived at by the same
procedure as that employed by Krogh (1938), i. e. by extrapola-
tion back to the time 0. Strictly speaking, of course, this procedure
is not permissible, as at no time do the curves for the concentration
fall become rectilinear; but as after the first rapid fall they do
assume a lesser curved course, one can by means of the extra-
polation form an idea, if a rough one, of the concentration in
■^vhich the substance -would be found if the “rapid phase” of the
distribution had taken place instantaneously. Of course, the
volume calculated from this would not represent the final distri-
bution volume, and at most it can only be of interest as a basis
for comparison with other series of experiments. The table clearly
illustrates how important it is to know the distribution time
when it is required to judge and compare results obtained. It -will
also be seen that the Aspersion in the results from rabbits, given
the same pre-experimental treatment, is moderate.

In experiment 12 a dehydration was intentionally produced:
the rabbit was first starved for 24 hours, whereafter the pylorus

IN0UN AS AN

vras

INDICATOR ROB THE EXTRACELWLAR SPACE.

ligated and tlie animal was left in a cage for 30 lioum w
• *L_- -.^Mrvr.4-i'n/T vrro f t>T rvr Ifind or of coDropliagv . A1

25
itli

5 lisated and tlie animai was ... .«

no opportunity of ingesting water or food ^ .

the contents of tlie stomacli ndth 570 mg. chloride had been with-
drawn through a tube, the experiment was performed the

usual technique. The values of the various volumes ot distribution

lie distinctly below the normal average. _ , ■ j

Pig. 1 and 2 show that at the end of the experimental period
the inulin and SCN" concentrations are still falling. The question
then arising is whether this is hecausc the distribution is not com-
pleted at this juncture or the substances arc eliminated in the
organism. In some Bases the experiments were therefore extended
over a much longer period. Here the practical difficulty was
encountered that the nephrectomized rabbits could generally
not be kept alive for more than 7 — ^10 hours. In the experiment
(No. 9) carried out over the longest periodj 12 hours, it was found
that the SON" concentration had practically become quiescent;
but at the end of the experiment there was still a slight fall in the
inulin concentration. After 12 hours the inulin was found to ho
distributed in 18 per cent., the SCN~ in 30.5 per cent, of the body
weight.

As to an eventual elimination of inulin, attention was given
especially to two possibilities; an extra-renal excretion and an
absorption into the reticulo-endothelial sj^stem. H.vv^YOOD and
Hobeb (1937) found that inulin could be excreted through the
isolated, artificially perfused liver of a frog. Therefore in a few
cases after the termination of the experimental period the gall-
bladder bile was analyzed for inulin (after removing interfering
pigments by means of carbon absorption in acetic acid solution).
It ivas found that the content of chromogen substances (by’ the
usual resorcin reaction) was only insignificant. In order to examine
the possibility of absorption into the reticulo-endothelial system
a couple of experiments were made in which the system was
the injection of a dye, Jacobsen and Pltoi’s method
( 45) of injecting 10 ml. 1 % trypane blue intravenously three

mib^d r ^'^tual experiment was

LnTe in th7"^"" if". - domonstrahl

IB hg. 3 is also shown the mult of o mlpUle inlmion. In tin

POOL KRUH0FFER.

experiment, after most of the distribution had taken place (295
minutes), 120 ml, of the previously described sulphate solution
was infused intravenously in the course of 26 minutes. As will be
seen from the figure, this brought about an increase of the volumes
of distribution for inulin and SCN". As regards the former the
increase is only very slightly larger than the infused volume.

ISOHN AS AS INDICATOR FOR TUB EXTRAOELDULAR SPACE. 27

whereas the increase of the volume of distribution for SON
is somewhat larger. The sulphate infusion moreover involves, a
dilution of the chloride in the organism, and, as already stated,
this permits of a calculation of the volume of dilutable chlondes;
the results of such calculations are also included in the figure.
As the diffusion coefficients of chloride ions and tliiocyaiiate ions
are of the same order, these volumes are best comparable with
the distribution volumes for SCN~, viz. by comparing a distri-
bution volume for SCN“ found after a certain period after the
SON" injection with a volume of dilutable chlorides after a cor-
responding period subsequent to the sulphate inhision. The volume
of dilutable chlorides is then found to be a good deal smaller than
the corresponding distribution volume for SON — but signifi-
cantly larger than any distribution volume for inulin.

In two similar experiments (Nos. 10 and 11) parallel results
were obtained. In Experiment 10 the volume of dilutable chlorides
were calculated 30, 90 and 170 minutes after terminating the
sulphate infusion (120 ml.) they were found to be 20.9, 22.6 and
23.6 per cent, of the body weight, and in experiment 11: 30, CO
and 120 minutes after terminating the sulphate infusion (120 ml.)
they were found to be 20.9, 22.6 and 23.6 per cent, of the body
weight.

In two experiments the ’peritoneal fluid was analyzed at the
end of the experiment immediately after the final blood sample
was drawn. The results were as shown in table 2, In this table the
concentrations theoretically to be expected in ultra-filtrates of
plasma are calculated as follows: for inulin, by multiplying the

pl.3maco,iceatt.tioabyl^„afo,g(;j,-.(cxpies»ea as NaSON)
by mnitiplying by

From the table it will be seen that in both cases the inulin

mSv! "" peritoneal fluid lies a few % above the theo-
reticaUy expectable concentration for an ultra-filtrate. The expla-

rOUL KRUn0FFER.

Table 2.

Distribution between plasma and peritoneal fluid.

Exp.

No.

Jlinutcs
after start
of exp.

JIatcrial

Plasma
cone, in
n>g %

n =

theoretic

ultrafilt.

b =
Cone,
found in
peri ton.

fluid
(mg %)

-100 =
a

found in
% of
thcoret.

Inulin

57.0

60.O

61.5

102.5 %

515

SCN" (exp. as

NaSCN)

25.0

27.7

25.0

88.5 %

Inulin

50.2

53.0

55.5

105 %

430

SON" (exp. as

NaSCN)

21.3

24.0

21.0

91.5 %

As previously stated, in two cases analyses were made on ten-
don tissue excised in the course of the experiments. In experiment
No. 14 (rabbit) the tendon was excised 150 minutes after com-
mencing, when the inulin concentration in the plasma water was
found to be = about 77 mg %; on analysis the tendon- tissue was
found to contain 17.8 mg % inulin, which corresponds to a con-
tent of about 23 ml. ultrafiltrate per 100 g. tendon. In the dog
experiment (No. 16) the results were similar; the tendons were
excised 130 minutes after commencing, at a time when the plasma
concentration of inulin was 70.2 mg % corresponding to a con-
centration in an ultrafiltrate calculated in the above mentioned
manner of about 73 mg %; at the same time 11 mg % was found
in the tendon tissue. The corresponding figiires for the SCN“
concentrations were about 21 mg % and 8. 3 mg %, which means
that in 100 g. tendon tissue there should be a content of 15 ml.
ultrafiltrate according to the inulin distribution ratio or of 40 ml.
according to the ratio of the thiocyanate ions. As the water con-
tent of the tendons is about 65 % most of which is extracellular
owing to the histological structure, it -will be seen that after the
period mentioned the inulin was not nearly distributed through-
out the extracellular space, whereas in the last experiment the
thiocyanate ions had attained to a much higher distribution
ratio. However, the tendon analyses embody not insignificant
sources of error, for which reason the results must be judged with
some reserve.

IffUWN AS AN INDICATOR FOR THE EXTRACELLDl.AR SPACE. 29

Discussion.

Before judging the results arrived at in this work it^ may he
of value to survey certain results obtained mth other indicator

substances by earlier workers, , »

A great part of the determinations of the “extracellular volume
has been based on the content of sodium and chloride, the postu-
late having been set forth that these substances were distributed
evenly and exclusively in the extracellular space.

However, recent publications have made the correctness of
this assumption very unlikely and in the following some of the
more significant arguments, which have been raised against the
theory that these two substances ate true indicators for the extra-
cellular volume, shall be summed up.

1) If the ratio m. cq. Na/m. eq. Cl in certain tissues is higher
than the corresponding ratio in an ultrafiltrate of plasma, both
sodium and chloride cannot be true indicators of the extracellular
volume of those tissues and most probably sodium gives values
that are too high. (There is a possibility, however, that chloride,
contrary to sodium, may not permeate into extracellular structures
(collagenous fibrillae, etc,).)

Moreover, if the ratio Na/Cl for another tissue of the same kind
of animal (or the same tissue under other conditions) is found
lower than in plasma ultrafiltrate it is most probable that both
substances give too high values for the extracellular volume in
the organism as a whole.

Harrison, Harrow and Yannet (1936) investigating the
electrolyte content of total bodies of rabbits, monkeys and dogs
found the Na/Cl ratio on an average 25 % higher than in plasma
ultrafiltrate. In bone and cartilage tissues the ratio Na/Cl has
repeatedly been demonstrated highly to exceed that of plasma
ultrafiltrate. (Harrison (193T), Iob and Swanson (1938), Har-
rison, Harrow and Yannet (1936).)

On the other hand, for certain other tissues Na/Cl has been
iMnd lower than m a plasma ultrafiltrate; in connective tissue

Hastings - 1938), in gastric and
mal mnwsa (hlANERv and Hastings — 1939), in liver

(Y™t and HimRow - 1910) and in immature rat uteri (Tal-
bot, Lowry and Astwood ■ 1940) ^

In striated muscle Hastings and Eichelberger ( 1937 )

POUL KRUHBFfEIt.

always found an excess of sodium over cUoride, while IMiller
and Dabbow (1940) found Na/Cl capable of varying: at low potas-
sium content of the muscles sodium was in excess of chloride
whereas at high potassium contents the reverse was the case.
(Heppel (1939) observed that muscle potassium could be replaced
by sodium at low plasma potassium concentrations).

2) The permeability of certain cells which may easily be ex-
amined may serve as a guide.

It is well known that erythrocytes contain chloride (and small
amounts of sodium), and that both the glandular cells of the in-
testinal tract and the tubuli cells of the kidneys are permeable to
these ions. Thus bj-- analogy it seems probable that other cell
membranes are also permeable to these ions.

3) If in one or other tissue the available space calculated for
a certain substance is obviously in disagreement with that deter-
mined by histological investigation this substance cannot be used
as an indicator for the extracellular space in that tissue. For
example, when SIaneby and Hastings (1939) found that calcu-
lated on the chloride content and expressed as a percentage of
the blood-free and fat-free tissue the extracellular phase amounts
to 22 % in liver, 32 % in spleen and 50 % in kidney (rabbits)
these values seem incompatible with the histological structure
of these tissues, even if for the kidney one makes a suitable correc-
tion for chloride contained in the tubulus urine. The same applies
to sodium. Analogous figures for cats’ organs will be found in a
publication by Ambebson, Nash, Miilder and Binns (1938).

4) As a true indicator for the extracellular volume must be
found in that volume exclusively and moreover only in a free state,
the concentration of such a substance in the tissues should vary
proportionally with the concentration in plasma (when sufficient
time for complete distribution is allowed for). If in certain tissues
a smaller or larger fraction of the indicator does not readily fol-
low the variations in the plasma concentration, that indicator
will beyond doubt “indicate” a volume larger than the extra-
cellular volume in these tissues.

Here the fine plasmapheresis experiments ot the last-named
four authors are particularly illustrative. By infusing sulphate
solution i. v. while simultaneously depleting for blood a gradual
fall in plasma chloride to very low values was produced.
WTiereas the chloride in some tissues (heart, liver, kidneys) fell
proportionately with plasma chloride, other tissues (stomach,

INtJLIN AS AN INDIOATOE FOB THE EXTBACELOTLAB SPACE. 31

spleen, pancreas, tendons, skin
system) 'were found to contain i

ride which was not removed. , v ■ i

Yannet and Dabeow (1940) found that muscles and liver also

contained small amounts of not-readily diffusible chloride.

(As pointed out by IVIaneby, Danieesen and Hastings (1938)
such nonfreely diffusible electrolytes need not be localized intra-
cellularly, they may also be bound to extracellular structures (in
larger amounts than their water contents can account for).)

Thus it seems reasonable to conclude that the available spaces
calculated from sodium or chloride naturally contained, in the
organism and individual tissues are larger than the actual extra-
cellular volume. Therefore, it is also to be supposed that the
volumes determined by administration of these substances (even-
tually as isotopes), are too large.

Logically, if the available spaces calculated from experiments
with other indicators turn out to be larger or equal to those cal-
culated on the basis of sodium or chloride they must be con-
sidered larger than the true extracellular volume.

Apart from Na'^' and Cl“, the substances which have been
proposed for the determination of the, extracellular volume are
chiefly SON" and Br~; determinations have also been published
with SO 4 , Sucrose and Mg'*"*'.

Eegarding bromide, as Weie and Hastings have shown that

and especially the central nervous
L smaller or larger fractions of cblo-

it is distributed in exactly the same manner as chloride in all tis-
sues (except the central nervous system) we should expect an
available space identical "with that of chloride.

Among the others, judging from the permeability found in
erythrocytes, glandular cells of the intestinal tract and the tubular
cells of the kidneys, should be expected to be a rather

good indicator while sucrose (like inulin) should be ideal.

In table 3 a nupiber of results from the literature are sum-
marized. It contains both available spaces calculated from the
content of Na+ and Cl“ in the organism and volumes of distri-
bution calculated from substances administered (usually intra-
venously and always in a single dose).

In the case of administered substances the results are not
directly comparable as the distribution volume depends on the
time elapsed after the injection and on the diffusion rate of the

af w alw "

y ffercnt and not eijually correct formnlae and finolly it

POUL KRUH0FFER.

Table 3.

Detenninalions of the volumes of distribution for different materials = N
calculated J) from the content or 2) from a single dose after a distribution
time — t. (cited from the literature). Besulis in % of body weight.

Man (healthy)

Dog

Babbit

Cat

ci-

from content

26—28 %»

25, 28 and
28 %'

ca. 23 %’
ca. 32 %•

Na+

from content

32—37 %>

=‘Na

from dose

—

28 %>
t = 3 min.
29 %‘

t = not stated

—

SCN-
from dose

20—28

t = 60min.or more
ca. 19 %•
t = ca. 30 min.*

ca. 30 %’
t = 60 min.

ca. 26 %•
t = 0 min.’

ca. 29 %*
t = 30 min.

Br-

from dose

23—31 %•»
t = 17s — 24 hours

24—38 %’»
t = 1 — 30 hours
21—30 %“
t = 6 — 497 min.

—

SO —
from dose

20-28 %»
t = 60min.ormore

21—32 %“
t = 42 — 79 min.

—

Sucrose
from dose

17—20 %«
t = 90 — 120 min.

1) 17.7—20.7 %»
t = 1 — 6 hours

2) 19.7—26.4 %”
t = 1 — 6 hours

—

Mg++
from dose

19—26 %“
t = 42 — 12 min.

—

* Haerison, Darkow and Yarnet (1936). * Asiberson, Nash, Mttlder and
Bikns (1938), calculated from results in table 1 and protocol in the following way:

= 565 m. ’ Hahr, Hevesy and Babbe (1939). * SIakery
4.1 • 100/9o • 100/9o '

and Bale (1941). ‘ Laihetes, Boerdillok and KuKonorrER (1936). ‘ Stewart
and Eourke (1941). ’’ Surdermarn and Dohar (1941). ' Lards, Cdttiro and
Larsor (1940). • ICrooh, A. (1938), tho plasma concentration at t = 0 was deter-
mined by extrapolation. Brodie, Brard and Leshir (1939). “ Wirkler
and Sjtrrn (1938). ” Sjiith, Wirkler and Schwartz (1939). ” Keith and
Power (1937).

inoun as an indicator for the extraceldbdar space. 33

must te remembered that determinations on falling plasma con-
centrations are apt to give too bigb values. ^ ^

Generally speaking, it is however seen from the table that there
are scarcely any of the other substances that show a volume of
distribution definitely lower than chloride, though sulphate and
sucrose have a slight tendency in that direction, thus it should
be supposed that all are distributed beyond the extracellular

space. _ , 1 1 • r n

The present work shows that in the rabbit and dog inulin be-
comes distributed in a perceptibly smaller volume than SON
(and Cr), and it is a well-known fact that inulin does not per-
meate the erythrocytes and cannot be assumed to pass into the
tubulus cells of the kidneys. This work has also shown that it
does not permeate liver cells (rabbit).

The question now is whether the available volume for inulin
is a true measure of the extracellular space?

The answer cannot be given definitely in the affirmative, but
it is justifiable to suppose that it is.

On the one hand, the available volume for inulin is scarcely
too large to represent the extracellular space; it is distinctly
smaller than the other “volumes” which several authors consider
should represent the extracellular space. Moreover, inulin's higher
molecular weight makes it very probable a priori that it does not
permeate the cells of the organism.

On the other hand, it appears from the present work that inulin
is cafoble of distributing itself regularly between plasma water
and interstitial fluid — see the analyses of peritoneal fluid. There-
fore it would be unnatural to suppose that the inulin does not
distribute itself throughout the whole of that part of the extra-
cellular fluid which is found extrastructurally.

However, there is a possibility that the inulin does not make
its way into the intrastructural part of the extracellular space,

by which is meant the fluid in structures such as collagenous
fibrillae etc.

The above mentioned investigations on the volume of distri-
bution of inulin in tendons might be expected to elucidate tliis
question.

As it will be remembered, in two experiments it was found that
after 150 min. respectively 130 min., inuUn was distributed in
23 respectively 15 ml of the 65 ml total water contents in 100 g
tendon tissue. Thus in both cases, in the time allowed for Zl

3 i60215. Acta pltys. Scandinav, Vol.ll,

Poni, KRUH0FFER.

bution, inulin had by no means spread into the whole of the
extracellular space in tendons; on the other hand, at any rate in
the former case, inulin was distributed in a larger volume than
the extrastructural water in tendon tissue can be imagined to
constitute.

Conseq,uently it seems justifiably to assume that the inulin, if
only it is given sufficient time, %vill distribute itself over the whole
extracellular space of the tendons. A long time will be required,
however, owing to the sparse vascularization of tendon tissue.

Even if inulin must be assumed to be a correct indicator-
substance for the extracellular space, its low diffusion rate will
be a very considerable handicap to its practical application for
determining the extracellular space in total organism as well as
in individual tissues (especially in those vuth sparse vasculariza-
tion). In the case of non-nephrectomized animals it can only be
employed for determining the extracellular space in individual
tissues after protracted, continuous inulin infusions.

As vdll have been seen, the tliiocyanate ions distribute them-
selves in a volume far exceeding what can theoretically be cal-
culated from the distribution curve of inulin. Provided that
inulin distributes itself only in the extracellular space and through-
out the whole of it, it -will therefore be necessary to assume that
SCN“ also distributes itself into a part of the intracellular fluid
of the organism. Eor instance, in the experiment illustrated by
fig. 1, after only 41 minutes SCN“ was found to be distributed
in a volume 185 ml. larger than could theoretically be expected.
If the erythrocyte volume of this rabbit is put at 80 ml., the
water phase in these 80 ml. will be about 50 ml. Thus in addition
to a (well known) distribution in the eryt^ocyte fluid there must
also be a penetration by SON" into other cells of the organism.

Summary.

A number of results are quoted from the literature, showing
that the available volumes for Cl“ and Na+ are larger than the
extracellular volume; the author gives a tabulated summary of
literatme references to the size of the available volume for these
substances and the volumes of distribution for these and SON",
Br~, SO 4 , sucrose and Mg"'”*'.

In 13 nephrectomized rabbits and one nephrectomized dog the
decline in the concentrations of inulin and SCN“ in plasma

INCLIN AS AN INDICATOR FOR THE EXTRACELWLAR SPACE. 35
was traced after intravenous injection of a single dose of inulin +
NaSCN.

It rvas found that: r

1) The experimentally determined curve of distribution tor
SCN~ (the volumes of distribution plotted against time) was
found to be placed well abo^m a theoretical curve of distribution
for SON", constructed from the assumptions that inulin and
SON are finally distributed in the same volume and that this
distribution proceeded exclusively by diffusion. The quantity
of fluid in the er 3 rthrocytes is entirely insufficient to explain this
discordance; it its presumable that SON- penetrates other cells
of the organism.

2) The distribution after a very long experimental period
(12 hours) suggests that the ultimate distribution in rabbits will
be such that inulin distributes itself in slightly less than 20 %
of the body weight and SON — in a little over 30 %,

3) In a few of the experiments, when the distribution of SON"
and inulin was well advanced, 100 — 120 ml. isotonic sulphate
.solution was infused into the rabbits. From the resulting dilution
of plasma-chloride the author calculated the “volume of dilutable
chlorides”; at a certain time after the sulphate infusion it was
found to be less than the volume of distribution for SCN“ at
the same period of time after the NaSCN injection.

4) Whereas the available spaces for SON" and Cl“ must be
supposed to be larger than the extracellular space, inulin must be
assumed to distribute itself in a volume that is equal to the
extracellular space. Tests showed that inulin is not excreted in
rabbit bile, and experiments with blockade of the reticulo-endothc-
lial system seem to indicate that inulin is not stored' by this
system.

5) Inulin s slow distribution, especially in sparsely vascula-.
rized tissue such as tendons, is, however, a serious drawback to
its practical application for the determination of the extracellular
space, even in nephrectomized animals.

ivererences.

Amberson, W E., T. P. Nash, A. G. Mulder, and D. Binns Aracr
J. Physiol. 1938. 122. 22i. ’

mL’tm. Anderson, Amer. J. Digest. Dis. Nutrit.

36 POUL KRUH0FFER.

Hahn, L. A., G. C. Hevesy, and 0. H. Rabbe, Biocbem. J. 1939. 33
1549.

Harrison, H. E., D. C. Harrow, and H. Yannet, J. biol. Chem.
1936. 113. 515.

Harrison, H. E., Ibidem. 1937. 120. 457.

Hastings, A. B., and L. Eichelberger, Ibidem 1937. 117. 73.
HAvemann, R., Biocbem. Z. 1940. 306. 224.

Haywood, C., and R. Hober, J. Cellul. Comp. Pbysiol. 1937. 10. 305.
Heppel, L. a., Amer. J. Pbysiol. 1939. 127. 385.
loB, V., and 'W. "W. Swanson, J. biol. Cbem. 1938. 122. 485.
Jacobsen, E., and C. Plum, Acta physiol, scand. 1943. 5. 1.

Keith, M. M., and A. H. Power, Amer. J. Physiol. 1937. 120. 203.
Keys, A., J. biol. Chem. 1937. 119. 389.

Krogh, a., Acta med. scand. 1938. Suppl. 90. 9.

Kruhopfer, P., Acta pbysiol. scand. 1946. 11. 1. Ibidem. 1946. 11. 37.
Lands, A. M., R. A. Cutting, and P. S. Larson, Amer. J. Physiol.
1940. 130. 421.

Lavietes, P. H., j. Bourdillon, and K. A. Klingboffer, j. clin. Invest.
1936. 16. 261.

Manery, j. F., I. S. Danielson, A. B. Hastings, J. biol. Cbem. 1938.
121. 359.

Manery", j. F., and A. B. Hastings, Ibidem. 1939. 127. 657.
Manery, J. F., and W. F. Bale, Amer. J. Physiol. 1941. 132. 215.
Maurer, F. W., Ibidem. 1938. 121. 546. .

Miller, H. C., and D. C. Harrow, Ibidem. 1940. 130. 747.

Northrop, J. H., and M. L. Anson, J. Gen. Physiol. 1929. 12. 543.-
Peters, J. P., Body Water. Baltimore 1935.

Rosenbauji, j. D., and P. Lavietes, J. biol. Chem. 1939. 131. 663.
Sjhth, P. K., a. W. Winkler, and B. M. Schwartz, Ibidem 1939.
129. 51.

SoMOGYi, M., Ibidem. 1930. 86. 655.

Stewart, J. D., and G. M. Rourke, J. Lab. Clin. Med. 1941. 26. 1383.
SuNDERMANN, F. AV., and F. C. Dohan, Amer. J. Physiol. 1941. 132. 418.
Talbot, N. B., 0. H. Loury, and E. B. Astwood, J. biol. Chem. 1940.
132. 1.

Weir, E. G., and A. B. Hastings, Ibidem. 1939. 129. 547.

Winkler, A. AA''., and P. K. Smith, Ibidem. 1938. 121. 589.
Yannet, H., and D. C. Harrow, Ibidem. 1940. 131. 721.

Medical Physiology, University of Copenhagen.

From tli6 InstltutG of

Tlie Significance of DifCnslon anfi CoiiTCction
for tlie Distriliution of Sointcs in
the Interstitial Space.

POUL KRUH0FFER.

Received 10 Ociolicr 1945.

According to Starling’s -wenknown theory the balance between
the plasma fluid and the fluid in tlie interstitial spaces is governed
by the following two factors: 1) the difference between tlie hydro-
static pressure within the capillaries and in the tissues, 2) the
effective colloid-osmotic pressure of the plasma.

If 1 exceeds 2 in some capillaries or in the arterial end of the
capillaries while 2 exceeds 1 in other capillaries or in the venous
end of the capillaries the conditions for a flow: capillaries — in-
terstitial spaces — capillaries are fulfilled. Such a flow must
inevitable carry with it solutes for wbicb the capillary wall i.s
permeable, i. e. involve a convection of suck solutes in the inter-
stitial spaces.

However, different views have been lield on the question whether
the conditions for such a "convection-flow” are fulfilled in the or-
ganism.

ScHADE in 1927 sums up the results obtained by himself and
his collaborators in experiments on models of artificial capillaries
concluding that these experiments make it probable that in the
natural capillaries a flow of fluid out of the arterial end and into
the venous end of the capillaries occurs.

Judging from the measurements of the capillary blood pressure
and the colloid-osmotic pressure of the plasma then in hand,
IvROGH in 1929 considers such a flow unlikely.

POUL KROU0FFER.

Nevertheless, the follo'w’ing, later experiments support the as-
sumption that conditions for such a flow are fulfilled; Landis
(1930 a) on the capillary blood pressure in the mesentery of
mammals, Landis (1930 b) on the capillary blood pressure in
human skin and Krogh, Landis and Turner’s (1932) pressure
plethysmographic determinations of the variations in tissue vol-
ume accompanying 1) increase in venous pressure and 2) in-
crease in colloid-osmotic pressure of jilasma.

This is also the conclusion drawn by Peters (1935) in his mo-
nograph "Body Water”, p. 54.

Thus at present it seems justifiable to assume that such a flow
of fluid through the interstitial spaces — "a paracapillary circula-
tion” — occurs, inevitably involving a convection of capillary
permeable solutes.

It is, however, a question whether this convection is significant
for the to- and fro-exchange of solutes between blood and in-
terstitial fluid and for their further distribution through the
interstitial spaces or whether this exchange is mainly due to dif-
fusion. This question, which apparently has not previously been
tested experimentally is the main problem of the present paper.

Principles Underlying Ihe Experimental
Procedure Employed.

In an earlier paper (Kruhoffer 1945 b) a number of argu-
ments were presented in favour of the assumption that inulin is
distributed in and exclusively in the extracellular space. The
amount injected

formula; : — was used for the calcula-

concentration in plasma water

tion of the volume in which inulin is distributed at any moment
after the injection. The volume calculated according to this
formula is named “volume of distribution” .

The volume of distribution comprises plasma water, at first a
part of it but quickly the -whole of it. That part of the volume of
distribution which is situated extravascularly will in the following
be denominated “volume of interstitial distribution” .

The model shown in Fig. 1 may be a help to visualize the effect
of convection respectively diffusion on the curves of distribution
of two non-electrolytes possessing different rates of diffusion ns for
instance inulin and sucrose. The model has two compartments

distribution of solutes in the interstitial space. 39

diffusion coefficient for sucrose
inuUn and the ratio; coeffreient for inulin"

struct a theoretical curve of interstitial distribution for the su-
crose. The interstitial volume of distribution for sucrose at the
time t -mil in this case be equal to the determined interstitial
volume of distribution for inulin at the time t • n.

Then suppose I filled wth ^vatcr and two apertures in the
membrane through which water is streaming in and out from P
ouing to the stirring of the water. If this flow of fluid is infinitely
rapid in proportion to the rates of diffusion of the two solutes the
distribution is effected solely by convection and their experi-
mentally determined curves of interstitial distribution will be
identical.

Finally suppose that the distribution of the two solutes is
effected by convection as well as diffusion. Then the experimen-
tally determined curve of interstitial distribution for sucrose will
be situated beWeen the experimentally determined curve of in-
terstitial distribution of inulin and the theoretical curve of in-
terstitial distribution of sucrose herefrom calculated.

The reason why inulin and sucrose in the above considerations
have been chosen as instances is, that these two nonelectrolytes
are suitable for analogous animal experiments as they possess the
lollowing properties;

1) When distribution is complete they are distributed in the
same (extracellular) volume.

POCI. KRUH0FFER.

2) Their rates of diffusion differ considerably.

3) The elimination of both can be abolished by nephrectomy.

Two preliminary experiments have shown that sucrose 5 hours

after injection in nephrectomized rabbits was distributed over
respectively 17.5 and 19 % of the body weight. This result is in
good agreement -with the final volume of distribution for inulin
found in earlier experiments (Kruhoffer 1945 b).

In the animal body the distribution in plasma from a practical
point of view is brought about exclusively by convection, but
contrary to our model in which the mixing occurs with infinite
rapidity, the mixing takes some time involving that a certain
diffusion into the interstitial space will have taken place before
the distribution in the plasma is complete. For this reason deter-
minations of the volume of interstitial distribution in the animal
body are not quite so simple as in our model.

By injecting the dye T 1824 simultaneously with inulin and
sucrose it is possible to determine the volume of plasma water
with which the injected solution at any given time has been mixed
(the mixing volume). The interstitial volume of distribution for
inulin (respectively sucrose) is then calculated as; volume of dis-
tribution for inulin at a given time minus the mixing volume at
the same time.

Technique.

A technique corresponding to that published in a previous paper
(Kruhoffer 1945 b) has been employed.

In the course of 15 seconds a single dose of inulin -{- sucrose -}- Tijji
was injected intravenously on nephrectomized rabbits (about 3 kg)
in amytal anaesthesia. 200 mg of inulin -f 350 — 400 mg of sucrose
-|-2 mg Tie:i dissolved in 10 ml 0.8 % NaCl were injected.

To reduce the error on the determination of sucrose (see below)
the plasma glucose concentration was kept low during the experimen-
tal period by means of an intramuscular injection of 3 — 4 i. tJ. of in-
sulin about one hour before tbe beginning of the experimental period.

In plasma from blood stabilized with heparin the concentrations of
inulin, sucrose and glucose were determined.

Ti 82 ,: The dilution of this dye was determined at ditferent points
of time after the injection by means of a Havemaxn’s (1940) photo-
electric colorimeter. 0.5 ml of plasma from arterial blood was diluted
with 4 ml of a O.o % NaCl-solution. The readings were made in a
"Kleinkuvette” with a layer thickness of 20 mm using a yellow filter
(GGi). For zero readings a solution containing 0.5 ml plasma drawn
before the injection and 4 ml 0.9 % NaCl-solution was employed.

DISTKIBUTION of fOLUTES IK T»E IKTEBSTlTiAL fFACH.

Tho deirree of dilution ^VBS rend from a xlcT-

1 nf thi-: cur\'e different ddiilions of 1 mi ni O.o

;‘.w» "" '” '’"■

samples.

degree of
dilution

Tbe curve Fig. 2 sliow.* the variation? in th'’ dej;r»'e of dilution of
T,n, during an ox])eriinert1 . Kslrapolfttion to rero time hit? heesi tn-sde
according to the, tnodus opvrandi now prevalent iti plafina voliitne de-
terminations, only n.'f the ordinate the degree of dilution wits Kuhsfi-
tuted for the concent rat ion of dye. If mi-ving h.ad oernrred in«t.nnt.anr-
ously the degree of dilution at ?erf> time would hnve heen 10.«, its-
creasing from this point following the str.tighl line, the incline of which
denote.? the extravaseidar los? of Tj,.,. Hence, the volume of pl-aMnu

water in the cjq>erin)ont pre.^ented was; ]()• 11).? • ml.

The mixing volmnc for in,«tanr« at time ,'t jninutes wjw;

10 - lO.s*

th'i ^ a
100 'h.

Dcterim'natioii of tindiri ond aHcrose in plasma containing glucose.
As all methods for determination of .sucrose provett to he Hcn.sitivc
also to inulin and glucose it was necc.Msary to work out- n special tech-
nique. This technique i.s based on the prcvioii-sly publiHliml colon-
metric method for determination of inulin tKiiunomin H'lf) a) in
which the colour reaction on heating for fit) minutes to 100'' C with
resorcin dissolved in ethylnlcohol and hydrochloric acid is ninde tieo of,
in this reaction glucose and mjcro.se gh-c rise to a development of
me same yellow colour ns inulin, Imt the intonsit v of colour developfsl

POUL KKUHDFPER.

differs for the same weight of the three substances as will be seen from
table 1.

Table 1.

Glucose-

cquiv. to

Thus, 1 mg

Sucrose-

cquiv-. to

Thus, 1 mg 1

cone.

inulin

inulin is equiv.

cone.

inulin

inulin is cquiv. 1

(mg %)

to mg glucose

(mg %)

to mg sucrose |

8.125

O.CG

12.3

3.0

1.70

1.7C5 1

16.25

1.27

12.8

4.5

2.53

1.780 !

24.375

1.92

12.7

6.0

3.38

1.775 j

32.5

2.58

12.G

9.0

1.785 i

5.11

12.7

12.0

6.74

1.780 i

18.0

.lO.lG

1.775 i

From table 1 it can be deduced that in respect to colour develop-
ment:

1 mg of innlin is on an average equivalent to 12.05 mg of glucose.

1 mg of inulin is on an average equivalent to 1.78 mg of sucrose.

The determination of sucrose in solutions containing inulin and
glucose is based on the following three figures;

1. Total colour development expressed as mg % inulin.

2. Colour development after fermentation i. e. inulin concentration.

3. Reduction power determined after Hagedorn and Jen'SEX (1923)
which gives the glucose concentration as neither inulin nor sucrose re-
duces KjFeCNo. From the glucose concentration the colour dc\elop-
ment originating from glucose is calculated by means of the factor
deduced from table 1 and expressed as mg % inulin.

By subtracting inulin plus glucose (expressed in the above mentioned
way) from the total colour development, the colour development
originating from sucrose is found. From this figure the concentration
of sucrose is calculated by means of the sucrose factor from table 1.
In this calculation allowance must be made for the loss of inulin which
takes place by the removal af glucose and sucrose by fermentation.
The recovery percentage by this procedure has experimentally been
found to be 94. The figure for the inulin concentration therefore must

be multiplied by

100

The recovery percentage by determination of the

total colour development and the reduction power on Somog}'i filtrates
is 100.

The analytic procedure described in detail in a premous paper (Kru-
HoFFER 1945 a) has been employed. Only to obtain a complete remo-
val of sucrose as well as glucose by fermentation a somewhat larger
amount of yeast (4 ml of a 10 % suspension for about 7 ml filtrate)
had to be used, hence the lower recovery % (94) for inulin.

The accuracy of the sucrose determinations is of course highest
when the concentrations of inulin and glucose are as low as possible.
As mentioned above, the glucose concentration was kept low during
the experimental period by means of in-sulin. The inulin concentration
must be kept within 35 to 90 mg % to obtain the highest possible ac-

DI3TIUBUTC0N OF SOLUTES IN THE INTERSTITIAI, SPACE. 43

curacy the iiiulin
determinations. Only
moderate sucrose con-
centrations can be used
to avoid undesirable
osmotic effects.

The following ex-
ample shows the accur-
acy obtained at the
concentrations used in
the e.vperiments; To a
plasma containing 103
mg % glucose inulin
was added to a con-
centration of 40 nig %
and sucrose to a con-
centration of 75 nig %.
In 5 determinations the
following concentra-
tions were found; Inulin
39.7; 39.D; 40.3; 40.i;
40.5 and sucrose: 73.!);
75.1; 73.0; 72.4; 72. i'.
At concentrations of
this order of magnitude
inulin apparently is
determined with an
accuracy of 1 — 2 %
and sucrose with an
accuracy of S — 4 %,

Beterminalion of the
ratio behceen the di/fu-
sion coeflicients for
sjtcrose and inulin. Tliis
determination was made
using the same tech-
nique and the same
diffusion chamber as
described in a previous
paper (KBUiioFFEn 1945
b). At 37° C the ratio

^sucrose t

Pj was found to

he 2.98. (Tlie concentra-
tions used in the inner
fluid were 1,500 mg %
inulin and 1,000 mg %
sucrose.)

Calculations: The vo-

VOUL KRDH0FFEK.

lumes of distribution for inulin and sucrose have been calculated
according to the formula:

amount injected
concentration in plasma water

The concentration in plasma water was calculated as plasma con-

ccntration — — .

Experimental Results.

In Fig. 3 A the results from an experiment are presented graphic-
ally. The interstitial volumes of distribution (ordinates) are
plotted against time (abscissa). The lowesfcurve (o — o) repre-
sents the experimentally determined curve of interstitial distri-
bution for inulin. The upper curve (x — x) is the theoretical cun^e
of interstitial distribution for sucrose calculated from the inulin

curve by multiplying the abscissae from this curve with —

. 6.98

The dots ( • ) which represent the experimentally determined
values for the interstitial volumes of distribution for sucrose are
seen to fall close to the theoretical curve.

In fig. 3 B and 3 C the results from two corresponding experi-
ments are shown. A fourth experiment, not pictured, gave quite
similar results.

Discussion.

From fig. 3, it is seen that the experimentally determined
values for the interstitial volume of distribution for sucrose on
the whole fall close to the theoretically calculated curve. In ac-
cordance wdth the considerations made in the first part of this
paper, it consequently must be assumed that diffusion dominates
the distribution of solutes in the interstitial spaces.

All the experiments have in common that the experimentally
determined volumes of interstitial distribution for sucrose in the
first part of the experiments (about 20 minutes) are larger than
the theoretically calculated values. In a simple, rigid system as
the model fig. 1 such results could never be obtained.

This discrepancy might be explained in one of the following
ways:

1) In the organism the ratio might be higher than that

found in the experiments with the^ diffusion chamber. Here three
phases must be considered: the capillary membranes, that part of

distribution of solutes in the interstitial space. 45

the interstitial fluid wHch is placed intrastructurally (vrithin the
collagenous fibres, etc.) and the remaining part placed extra-

structurally. , , . . t.

As regards the last named, all the molecules it contains must be

considered freely movable, and the idscosity must be almost the
same as that of pure -water. Consequently, in the extrastructurally
placed part of the interstitial volume the diffusion rates of the dif-
ferent solutes must be almost the same as those found in pure water.

Concerning the t-wo other phases, the capillary membranes and
the intrastructural part of the interstitial space, it must be pointed
out that both are built up of fixed molecules, which confer to
them a definite molecular pattern. Beyond doubt, the rate by
-(rhich solutes find their way through such a pattern will depend
on their molecular size, the pattern offering more resistance to the
progress of larger molecules. We have no definite conception of
the composition and idscosity of the fluids filling out the lacunae
in these patterns and thus we have no knowledge of the degree
to which they affect the rates of diffusion for inulin and sucrose.

Summing up, it seems most probable that the ratio

for

'iDuUn

the distribution within these two phases should be somewhat
higher than in the case of pure aqueous solutions.

Compared to the extrastruotural part of the interstitial volume,
the intrastructural part and the capillary membranes possess only
a small volume and further the intrastructural part is that most

remote from the capillaries. Therefore, I think a higher

r>.

inulin

ratio in the two structural phases cannot be the sole explanation
of the discrepancy.

2) A living animal is not a rigid system as the model fig. 1,
If the membrane separating P and I during the first part of
the experimental 'period were larger than later on, the experi-
mentally determined interstitial volumes of sucrose would be
somewhat larger than the corresponding calculated values until
complete distribution is attained. The same thing would happen
in t e h^ng animal if the number of open capillaries was larger
urmg the first part of the experiment than later on. A successive
c osure of capillaries during the experiment, caused by a devel-
oping post-operative shock does not seem improbable and further
j secluded that the rapid, intravenous injection of 10
nu tluid induces a temporary opening up of capillaries.

POUL KEUH0FFER.

Certainly too, the spontaneous variations in the number of open
capillaries may affect the total size of capillary membrane sep-
arating at any given moment plasma from interstitial fluid. How-
ever, as these variations must be supposed to occur at random
about a rather fixed mean value they can hardly affect the results
in a particular direction.

On the other side it is very difficult to estimate the net influ-
ence which the opening-closure intermittence of the capillaries may
have on the course of the distribution curve; because this influence
must be composed of more factors affecting the results in different
ways.

By calculating the results in the manner mentioned above a
systematic error is made. Owing to the more rapid distribution of
sucrose a smaller percentage of tliis than of inulin will be removed
in each blood sample drawn. Calculating the distribution volumes
from the total amounts injected will thus result in estimating all
volumes of distribution for inulin relatively too high (absolutely
both are calculated a little too high). However the errors introduced
in this way are only of limited consequence. (Yet eventual further
experiments with the technique here suggested ought to be per-
formed on larger experimental animals to make the blood samples
represent only a minute fraction of the total blood volume.)

Though the technique here suggested thus involves certain un-
controllable sources of error when it is employed on the living
animal, yet I think the above conclusion that diffusion completely
dominates the distribution of sucrose and inulin in the interstitial
spaces valid. That convection in some (quantitatively little im-
portant) tissues, particularly in tissues (like tendons, cornea etc.)
with poor vascularization (long ways of diffusion), may play a
certain role cannot of course be excluded. To decide whether this
is actually the case determinations of the rates of distribution
for inulin and sucrose in these particular tissues should be carried
out, keeping the plasma concentrations at a constant level through-
out the experimental period.

The diffusion rates of sucrose and particularly inulin are lower
than the diffusion rates of some materials normally transported
(glucose, urea etc.), consequently convection plays a still smaller
part in the distribution of these materials in the interstitial spaces.

Hinally the attention shall be drawn to the fact that from the
renal physiology we know cells which can apparently be supplied
with certain solutes by diffusion against a fluid stream; \dz. the

DISTRIBUTION OF SOLUTES IN THE INTERSTITIAL SPACE. 47

cells of the proximal tubules which are able to receive consider-
able amounts of phenol red, diodrast etc. from the capillaries in
spite of a fluid stream in the opposite direction through the inter-
stitial spaces. (Walker et al. (1941) demonstrated that large
amounts of water must be reabsorbed in tubuU proximales.)

Summary.

In experiments on nephrectomized rabbits inulin and sucrose
have been found to have at least very nearly the same definite
volume of distribution viz. slightly less than 20 % of the body
weight. Both materials must be assumed to distribute themselves
only in the extracellular space.

As sucrose diffuses about 3 times as rapidly as inulin, sucrose
is more suitable for determinations of the extracellular space and
must at present be considered the most appropriate substance for
such determinations on nephrectomized animals and individual
tissues, however complete distribution is not obtained till 4 — 6
hours after intravenous injection.

An experimental procedure is described rvhich makes it pos-
sible to decide whether convection or diffusion is the most im-
portant factor in the distribution of solutes in the interstitial
spaces. The method is based on determinations of the so called
curves of distribution for inulin and sucrose after simultaneous
intravenous injection on nephrectomized rabbits.

Brom the results of experiments of this kind it is concluded
that diffusion completely dominates the distribution.

Beferences.

Hagedorn, H. C., and B. N. Jensen, Biochem. Z. 1923. 135. 46.
Havemann, R., Ibidem 1940. 306. 224.

Keogh, A., The Anatomy and Physiology of Capillaries, New Haven.
1929. 2nd ed.'p. 29^

Keogh, A., E. M. Landis, and A. H. Tuenee, J. Clin. Invest., 1932.
11. 63.

Keuhgffee, P., Acta physiol, scand. 1946. 11. 1. Ibidem. 1946. 11. 16.
Landis, E. M., Am. J. Physiol. 1930 a. 93. 353.

Landis, E. M., Heart. 1930 b. 15. 209.

Peters, J. P., Body Water, Springfield 1935.

ScHADB, H., Brgebn. inn. Med. 1927.' 32. 425.

Walker, A. M., P. A. Bott, J . Oliver and M. C. Me Dowell, Amer.
J. Physiol. 1941. 133. P. 480.

Fysiologiska Institutionen, G. C. I., Stockholm.

On Acute Effects of Cigarette Smoking on Oxygen
Consumption, Pulse Rate, Breathing Rate
and Blood Pressure in Working
Organisms.

ANTS JUURUP and LEONID MUIDO.

Received 12 October 1945.

A considerable number of papers on the effect of smoking can
be found in the literature. The authors agree that in resting con-
ditions there is a rise in pulse rate and blood pressure as a result
of smoking. As to the effect on metabohc rate the opinions differ.
About the effects of tobacco on a working organism we were, un-
fortunately, unable to find any works in literature. From the
point of exercise physiology this, however, should be of particular
interest. It is known that the metabolic rate and other functions
show manifold rise in a working organism and that certain drugs
and hormones have quite different actions in a- resting and in a
working organism. Therefore, by making experiments on organ-
isms at the state of basic metabolism we cannot draw conclusions
to the working organism.

The following experiments, inspired by prof. E. H. Christen-
sen', try to give some experimental basis to suggestions of the
action of tobacco on working human organism.

Methods.

The experiments were made with the Krogh’s bicycle ergometer on
which it is possible to exactly determine the amount of work done. A
metronome was used for the timing of work controlling the rhythm of

AODTE EFFECTS OF OIGAKETTE SMOKING.

pedalling (60 turns per min.). The turns of the bicycle were recorded
by a cyclometer.

The rate of work varied in different experiments and has been re-
corded in the description of the experiments. Analyses of expired air
were made by Haldane’s apparatus for gas analysis. For the measuring
of blood pressure Kiva-Eocci’s sphygmomanometer was used. Pulse
rate was measured on A. carotis.

The determinations of respiratory functions, oxygen consumption,
pulse rate and blood pressure at the state of rest for the purpose of
control and comparison were made after 45 minutes’ full rest. A. J.,
one of the subjects, even slept in the laboratory. After rest the subjects
smoked, males two and females one cigarette, and 5 minutes after
smoking (to eliminate the effect of movements in connection with
smoking) new determinations were made. For smoking the same
brand of cigarettes (Hudson) was used every time.

Before the exercise test the subjects rested for at least 30 minutes
after which time pulse rate and blood pressure were measured. Then
the subjects smoked and five minutes after smoking a new' determina-
tion of pulse rate and blood pressure was made. Immediately exercise
test followed. These test exercises were made alternately, one day with
smoking, one day without, in order to eliminate the effects of training.
Before the experiments the subjects fasted and did not smoke at least
for 12 hours.

Experimental Eesults.

The first experiments were made on male subject L. M., 30
years, 77. G kg, 185 cm, moderate smoker (10 — 12 cigarettes daily)
during the time of 8. II. 45—12. VI. 45. The subject, an active
sportsman, was at the beginning of the experimental period un-
trained. The mean values and standard error of the mean e (M)
are brought in the tables.

From the experiments made at the state of basic metabolism
(Table 1) we can see that there is a marked difference in pulse

Subj. L. M.
Table 1.

Blood

Number
of exp.

Vent.

37°l/niin.

Breath.

rate

Oj 1/min.

Pulse

pressure

syst.

I no smoking

6.02

6.7

0.205

60.7

108.3

71.7

Rest-experiments. |

1 e (M)

± 0.23

± 0.8

± 0.002

± 1.7

+ 4.4

± 1.0

1 smoking

6.01

7.0

0.271

68.3

120.0

75.0

1 e (M)

± 0.20

± O.o

± 0.008

ib 2.4

± 2.9

± O.o

0 (M) = Standard error of the mean.
4 i60215. Acta phys. Scandinav. Vol. 11.

iiNTS J00RUP AND LEONID SlUIDO.

rates. After smoMng the pulse rate is faster than without smoking.
A definite influence can also be noticed between the corresponding
blood pressures; after smoking the blood pressure increases.

The first rate of work was 1,080 mkg per minute in 45 minutes
(Table 2). The respiratory functions have been determined every
15, pulse rate every 5 and blood pressure every 10 minutes. Def-
inite changes were obtained only in pulse rates. There is a rise
also in blood pressure which, however, is not statistically certain.

The next rate of work was 1,260 mkg per minute in 30 minutes
(Table 3). Kespiratory functions were checked here as well as in
following rates after every 10, pulse rate after 5 and blood pres-
sure after every 7 minutes. Here we can also see an increase only
in pulse rate, whereas other functions show no statistically signif-
icant changes.

Applying 1,440 mkg per minute in 20 minutes we see the same
changes (Table 4). Sometimes even the pulse rate shows a sta-
tistically uncertain increase, particularly at the 5th minute. At
the 20th minute we can again see a pulse increase of 18.7 beats
per minute, e (hi) being ± 5.2.

The last rate of work was 1,560 mkg per minute in 10 minutes
(Table 5), During these experiments the subject reached the limit
of his capacity. He found it difficult to keep pace and complained
over weariness which could also be ascertained objectively. Even
at this rate of work the effect of smoking can be recognised only
in changes of pulse rate.

At the end of these experiments pulse rate and blood pressure
were measured after 5 minutes’ rest and even here we can see
that after the experiments with smoking the pulse does not reach
the same low level after 5 minutes of rest as after the experi-
ments without smoking.

The next experiments were made with A. J., male subject, 29
years, 85 kg, 183 cm, light smoker (3 — 4 cigarettes daily), well-
trained.

In the state of basic metabolism we saw similar changes, there
was a definite rise in pulse rate while blood pressure showed
only a small rise which could not be regarded as statistically
certain.

Exercise experiments were made on him with 1,260 mkg per
minute in 30 minutes (5 experiments with and 5 without smoking).
1,440 mkg per minute in 20 minutes, 1,620 mkg per minute in 10
minutes and 1,800 mkg per minute in 10 minutes. With the last

Table 3.

ANTS JUURUP AND LEONID MDIDO.

.S o

s§

oPh

• O

C *-«

£ S

■5J

O ^

E 2
^ O

00 CO o ‘O

^-hS-h

CO O OO Cl

cd CO CO i-H

« O CO o

odidcdo

S-hS-h

o 1- *.0 »«
OOOr-^

S-h“-h

OO 00 o

o5oo3o

-« -H

•rji O rf< O

(MtHITOO

c:'

^ to'^

O o c o

CQ O

O S

C CO

20 min.
Pulso

151.3
± 2.1
159.5
± 3.8

»i4

wt to O w'>

fc-r:

COCOO’^

(*-(

»oP

S-h'S^i

■iS

»f:> 05 o o

f<5

C **5

03COkOO

ae.

Til'd

tH O

CO rt •f'

00*^0 03

»*

iSniS-H

o o «0 -<

o o o ©

030030

+1 -H

t- wo © ©

kf5 O O O

b 5

CO C5 WO

CO fs. C\J

o o

-H -H

b .

•2 »<
a ®

r}4 •}«

3t«

!2; o

ritf bO'^

o o c o

S !3

CO O

? s

C CQ

rest.

press.

.Ad

« h- cc -r

50 CO 06 03

Blood

4:3

116.3
± 2.8
116.0
± 2.9

0 © to «
0006’^

i P

.5 o

s-s

03
. 03

.£ "

71.3
± 2.9
74.0
± 2.0

'T3

00 0

w-S Cl © «

l> CO 06 pH

S-hS-h

.3 O

a-i

i ^1

E ^

•pH O

■SJ

^ D-
CO

D-'pHCDCO

1 -HCOlO

o O to o
0<NI> w

00 03 00 tH

S-hS-h

030030

-H -H

C5 wo 00

idooo

toe* e*
.s^

^ tc
O O c o

a s

00 o

o £

CS 03

‘i

(M) = standard error of the mean.

Table 4

acute effects of cioarette smoking

kid

diast.

O O O CO

»OOOC<3

•4^

CO CO CO

COr-teOCO

S-3

o(^

c "

■p ^

^ §

>o'^

^ o
O c5

b< 4 h

c a

-§ ^
S s

O O
00 tH CD CO

CO »o CO «>
Ot!OJ>CO
r* Ci

’-' -H

b. b- UO

CD tH ^ T—f

'^-H”4f

O O b> O

oo 3 t- 3 t 4

S-hS-H

OO b* b; •«#
^ ys

^ "H

coocoo

kfj CO CO b«

trfOlOO

C^<M-M<CO

,i!

a .

o.a

s a

0 (M) = Standard error of the mean.

ANTS JUURDP AND LEONID MUIDO.

Fig. 1. Effect of smoking on pulse rate at different amounts of irork. Subject

A. J. smoking no smoking. Each point represents the mean of 3

to 6 single determinations.

three rates of work 3 experiments were made with and 3 without
smoking. The results showed that ventilation and breathing rate
as well as the oxygen consumption are stable, showing no effect
of smoking, whereas pulse rate again showed a definite rise. Here
we could see once more that after 5 minutes the rest-pulse rate
still being under the effect of smoking did not reach the same
level as in other experiments made without smoking. Fig. 1 shows
the effect of smoking on pulse rate at different amounts of work.
Curve of the pulse rate after smoking runs at a higher level.

ACUTE EFFECTS OF CIGARETTE SMOKING.

With the subject G. B. male, 25 years, 73.9 kg, 171 cm, vrell
trained and non-smoker, only one series of experiments were
made, namely with 1,620 mkg per minute in 10 minutes, 3 ex-
periments were made with and 3 without smoking. Here roughly
similar results were obtained. The subject did not feel well after
smoking and was once compelled to finish the experiment for
that reason. In pulse rate we could see a big difference after 5
minutes (20 beats per minute, c (M) being ± 6.9), that however
diminished after 10 minutes (5 beats per minute, e (M) being
± 7.1) becoming statistically uncertain. The subject told by the
way that the first 5 minutes after smoking had been the most
strenuous, later on the driving had been easier.

The experiments made on female subject H. R., 28 years, 172
cm, 72 kg, light smoker (2 — i cigarettes daily) gave following
results:

At the state of basic metabolism the results showed correspond-
ence with the previous ones. There was a sure rise only in pulse
rate and blood pressure. After smoking the subject complained of
dizziness and nausea.

The first rate of work was 540 mkg per minute in 20 minutes
(Table 6). Pulse rate shows no changes. The subject admitted
that as she had not felt well at the prc\dous experiments (at basic
metabolism) she tried to smoke with caution this time. That could
also be seen from the rise in pulse rate at rest. At the state of
basic metabolism the rise in pulse rate was 13 beats, here only 8.

The fact that the pulse rate here does not give higher values
at work depends on the short duration of the effect of tobacco
in this subject. An experiment was made on her at the state of
basic metabolism: when after a 30 minutes’ rest she smoked one
cigarette her pulse rate showed the resting-state values already
at the 10th minute. Considering that the beginning of exercise
test was actually 5 minutes after the smoking we can understand
why the 5 minutes’ pulse at work (i, e. ten minutes after smoking)
does not show rise any more when compared with the pulse rate
without smoking.

At the next experiment, 720 mkg per minute in 20 minutes
(Table 7) the subject smoked 2 cigarettes instead of 1 at the
previous experiment, trying to deeply inhale the smoke. She had
nausea after each smoking, after two experiments she vomited
(there were three experiments with and three without smoking).
The results show a pronounced rise in ventilation and breathing

Table 0.

AKTS JUURT3P AKD LEONID MUIDO.

C5 COOtS

CO CV3 ^ <M

•-1C0C5O

C^COOtH

CO ,,co ,,

tHOtHO

-H -H

05
. 05

45> O
05 »-•

£ ^

O « o

t-5i-HC'3 03

C O

s 1

112.3
± 3.5
115.8
± 2.2

After

Pulse

78.0
±3.i

80.0
± 2.5

20 min.

148.0
±3.3

146.0
± l.G

. 03

-S "

B.?

O © W
OCOCOth

•rJ

C5 O
i-H O

syst.

O 03 Cl

i>cr5o3co

1 > t

15 min.
Pulse

o o «
CO eo t-5

Tj< ^

. 03

S 2

diast.

00 rt

i> ^ 00 t- 5
to to

-fl 41

r}< O
“i-i O

ayst.

150.0
±-2.l
156.3
± 6.2

« rt O «

-?• o « ©
rHOr^O

«5 •*<»-• »»

coot>o

-H -H

CO •J' »-* «
CD0I>0

-H -H

-H

“c- cr

S ^

0 o c o

OQ Q

g s

C CO

,£

bo'^

o o c o

s ^

w o

° s

C 03

o (M) = standard error of the moan.

Min. of tirnrk

ACUTE EFFECT? OF CIGAlinTTK PMOKINO

(.M) St.in'lnni error of tho mc:in,

ASTS JOORtIP AND LEONID MDIDO.

rate. Oxygen consumption is practically unchanged again. Pulse
rate shovrs no rise.

Increased ventilation and breathing rate at 720 mkg can be
taken as symptoms of acute nicotine poisoning. There were other
symptoms: increased secretion of saliva, nausea, vomiting, cold
sweat.

Lastly attempts were made to determine duration of the in-
fluence of smoking. Experiments were made on L. M. and A. J.
After a rest of 15 minutes they smoked 2 cigarettes each which
was followed by 30 minutes’ sitting and a 30 minutes’ lying.
After that time (t. e. 60 min. after smoking), they made exercise
test with 1,440 mkg per minute in 10 minutes. The results show
that after 60 minutes there was no influence whatever of the
smoking to be seen. After 45 minutes there was still a slight in-
fluence of smoking in L. M. The pulse rate was higher. In A. J.
the influence was still noticeable after 30 min.

Discussion.

As already mentioned we were unable to find any papers on the
effect of smoking on the working organism. Quite a number of
papers, however, could be found describing the influence of smok-
ing on the organism at rest.

Goodman (1914) claims that smoking increases blood pressure
by 10 — 30 mm. Hg. whereby the increase depends on the strength
of tobacco.

Fisher and Berry (1917) noted a rise in pulse rate together
with rise in blood pressure whereas the changes were not char-
acteristic.

Short and Johnson (1939) assert that changes in pulse rate,
blood pressure, skin temperature and blood sugar can be ex-
plained by increased secretion of adrenaline. By injecting 1 cc.
of 1 : 1,000 adrenaline solution they got similar results to
those after smoking. They consider it probable that this is the
stimulating influence of nicotine on the sympathico-adrenal
system.

Hiestand, Eamsey and Hale (1940) found that smoking in-
creases metabolism and pulse rate whereas breathing rate dimin-
ishes. These changes, however, do not correspond in all cases as
there is an increase in metabolism in 82 per cent, a decrease in

AC0TE EFFECTS OF CIGARETTE SMOKING.

13 per cent and no change in 5 per cent of all the subjects. There
is an increase in pulse rate in 72 per cent, a decrease in 26 per
cent and no change in 2.5 per cent of cases. The breathing rate
showed rise in 41 per cent, fall in 53.9 per cent and no change
in 5.1 per cent of cases.

Main (1941) finds that changes in pulse rate and blood pressure
last for 30 — 60 minutes after smoking.

Goddard and Voss (1942) assert that basic metabolism may-
be increased or decreased by smoking depending on the sensibility
of the person’s sympathico-adrenal system.

Steinmann and Voegeli (1942) examined changes in circula-
tion after their subjects had smoked and passed through a work-
test. They found that after work the minute-volume was in-
creased which was chiefly due to higher pulse rate whereas in the
stroke volume there was either a small increase or decrease.

All those statements correspond with the results as to the
changes in pulse rate and blood pressure under effect of smoking
obtained in this work. The changes in oxygen consumption, how-
ever, make an exception in so far as there was no increase under
the influence of smoking.

Although it is highly probable that smoking increases the secre-
tion of adrenaline in the suprarenals, -we can possibly not ascribe
all the changes that have been recorded to the influence of this
agent. The smoke of cigarettes, as we know contains nicotine as
well as carbon monoxide, HON, furfural, NHj, pyridine and its
derivatives like pyridyl piperidin, coUidin and nornicotine (an
isomer of nicotine), quinoline, phenols, carbon dioxide and some
volatile oils, which all in some way can influence the organism.

Christensen (1931) finds that extended training decreases rest
pulse as well as exercise pulse rate, both rates being lower in
trained than in untrained subjects. The pulse rate in trained
subjects reaches a certain level staying at this height until the
end of the work. In untrained subjects the pulse rate rises con-
tinually dining the work.

If we compare those data with the changes in pulse rate ob-
tained in this paper we can see that smoking has a contrary effect
to that of training on the pulse rate, -viz. it deteriorates the con-
dition.

akts juurtjp anp lponid muido.

Concltisions.

1) Smoking calls forth a rise in pulse rate at the state of rest
as well as at work.

2) Smoking has no effect on oxygen consumption.

3) It may be assumed that smoking does not affect ventilation
and breathing rate.

4) Rise in blood pressure as effect of smoking is probable.

5) At a certain oxygen consumption, pulse rate is higher under
the effect of smoking than no smoking.

6) The duration of the influence of smoking is 10 — 45 min.,
showing individual variations and depending on the amount of
tobacco smoked.

References.

Christexsex, E. H., Arbeitsphysiologie 1931. i. 453.

Fisher, G., and E. Berry, Physical Effects of smoking. New York
1917.

Goddard, V. R., and Y. G. Voss, J. Lab. Clin. Med. 1942. 27. 789.
Goodjiax, E. H., Blood pressure in medicine and surgery. Philadelphia
1914.

UiESTAXD, AV. A., J. Ramsey, and M. Hale, J. Lab. Clin. Med. 1940.
10. 1013.

Maix, R. j., Proc. Soc. exp. Biol. N. Y. 1941. 48. 495.

Short, J. J., and H. J. Johxsox, J. Lab. Clin. Med. 1939. 24. 590.
Steixmaxx, B., and H. A^'oegeli, Dtsch. Arch. Klin. Med. 1942. 189.
319.

From the Laboratory of Zoophj’siology, University of Copenbagen.

Amino Acids and Related Compounds in tlie
Haemoljinpli of Oryctes Jfasicornis
and Melolontlia Tnlgaris.

HANS H. USSING.

Ucceivod 13 October 1915.

One of the most characteristic features in the biochemistry of
insect haemolyraph i.s the high content of non-protein nitrogen
(DtA'AL, PoRTiER ct ah 1928). Indeed in most forms the greater
part of the osmotic pressure comes from amino acids or related
■compounds.

AMiereas amino nitrogen determinations on the haemolymph
of different insect species have been made by several investiga-
tors, our knowledge concerning the nature of the amino nitrogen
found is but scanty. Heller (1930) found that the amino nitrogen
of deproteinized Deilcphila-blood increased considerably on acid
hydrolysis, from which it was concluded that a considerable part
of the amino-N originated from peptides. He further found that
a considerable part of the amino-N could be precipitated by phos-
photungstic acid and by uranylacotate.

Florkin and Duciiateau (1942) were unable to detect argi-
nine, tryptophane, phenylalanine and cystine in the blood of
Dytiscus marginalis. 30 rag% histidine and 117 — 1G8 mg% tjTO-
sine were found, but this is quite insufficient to account for those
106 mg% amino-N, which the blood was shown to contain.

The question about the nature of the amino nitrogen of insect
blood thus invited to further study. Larvae of Oryctes nasicornis
and Melolontha vulgaris were chosen for the investigation. One
result, the isolation of asparagine from Melolontha blood has been
published in a note (Ussing 194.'5 a).

HANS H. HSSING.

One apparent implication of the high amino-N concentration
in insect blood is that these animals, especially the herbivorous
forms, have to take up amino acids from the intestinal contents
against a high concentration gradient, and the insects might pre-
sent a good object for the study of active amino acid uptake.

From the papers of Heller (1. c.) we learn, however, that a
considerable part of the amino-N originates from peptides in
Deilephila-blood and Florkin and Duchateau (1. c.) failed to
find several important amino acids in Dytiscus, so it could not
he taken for granted that a general mechanism for active amino
acid uptake is at work in the intestine of insects. It might well
be that the high amino-N content originated from one or some
few amino compounds. In the course of the present study it was
therefore decided to undertake some analyses on intestinal con-
tents in order to allow a comparison between the concentrations
of different amino acids in the blood and in the intestine.

Experimental.

1. Preparation of protein-free blood extracts.

An incision was made in the side just behind the head of the larvae.
The blood was allowed to fall directly into a measured amount of 20 %
trichloroacetic acid in a weighed vessel. "When a suitable amount of
blood — from several larvae — had been collected, the vessel was
weighed and the volume was made up to a multiplum of the blood
volume; then the mixture was stirred and left at room temperature
for one hour. The protein-free solution was filtered off into a measuring
cylinder and then freed from trichloroacetic acid by extraction with
ether in a continuous extractor. The ether was removed by aeration or
concentration in vacuo.

2. Separation of the blood-filtrate amino compounds in three

main firactions.

In order to get an idea of the distribution of the amino-N on the dif-
ferent amino-acid groups, blood filtrates were treated with phospho-
tungstic acid, which precipitates most peptides and the basic amino
acids. After removal of excess precipitant the filtrates were saturated
%vith Ba(0H)2 and the dicarbonic acids precipitated with about 5 vol.
alcohol. After removal of inorganic ions from precipitates and filtrates
as well, three fractions resulted which are in the following designed
as P^Y-fraction, Ba-fraction and monamino-acid fraction. The analysis
on blood from Melolontha may serve as an example:

AMINO ACIDS AND KELATBD COMPOUNDS.

The blood from 25 larvae, 7.1 g, was collected in 7 ml trichloroacetic
acid. "Water was added to the 70 ml mark and after mixing and standing
for one hour, the solution was filtered off. Filtrate 64 ml. A drop of
10 % sulfuric acid was added and trichloroacetic acid removed by
extraction with ether. The solution was concentrated in vacuo to a
small volume and transferred with water in several portions to a cy-
linder. Volume: 20 ml. 6 ml drawn for analyses: the rest made 2.6 %
as regards sulfuric acid and mixed rvith 1 g phosphotungstic acid, dis-
solved in a little hot water. After standing over night in the cold,
the precipitate was centrifuged and washed with dilute phosphotungstic
acid solution in 2.5 % sulfuric acid. The precipitate was suspended in
water and decomposed with an c.xcess of warm saturated Ba(OH )2 and
after removal of the precipitate and washing, the excess of Ba++ was
immediately removed by addition of H-SO^: P'W-fraction. The com-
bined filtrates from the phosphotungstic precipitate were freed from
HjSOi and excess precipitant by addition of warm saturated Ba (OH )3
and the precipitate washed three times with hot water. Tlie combined
filtrates, 25 ml, were saturated with Ba(OH)t and mixed with 125 ml
alcohol. After 18 hours in the cold, the precipitate was filtered off and
washed with 85 % alcohol. The precipitate was freed from Ba++, the
BaSOi being thoroughly washed with hot water: Ba-fraction. The
alcoholic filtrate was concentrated in vacuo until all alcohol had been
distilled off and then the Ba++ was removed quantitatively: mon-
amino acid fraction.

Table I.

Total N

Oryctes *•/,

Total blood (424)

PW-fraotion 181

Ba-fraction 22

Monamino acid fraction 250

llololontha •'/»

Total blood 402

PW-fraction 94

Ba-fraction 17

Monamino acid fraction IGO

Amino-N

mg%

316

180

280

129

Table I gives examples on the distribution of total N and
amino-N in blood-filtrate from Oryctes and Melolontba. Total
nitrogen is determined by micro-Kjeldal analysis and amino-
nitrogen (after removal of ammonia by distillation in vacuo at
pH 9) according to the modification of Folin’s colorimetric
method described by Ussing (1945 b).

The value for total nitrogen on untreated filtrate from Oryctes
is placed in brackets, because it refers to another blood sample
than that used for the other analyses.

IIAKS n. USSING.

It vrill be noticed, that the sum of the amino-N of the three
fractions from Oryctes gives 264 mg%, vrhereas the original
amount vras 316. In the case of Melolontha there is an even big-
ger loss of amino-N during the procedure. In order to locate the
loss, an experiment ^vas made on Oryctes blood where analyses
were performed on the filtrate from the P'Wrprecipitation, before
the Ba-alcohol treatment. The results are shown in table II. It
is seen that the calculated amino-N concentration of the PW-
fraction is 38 mg% higher than that found. Probably the loss
is at least partly due to adsorption of peptides to Ba-phospho-
tungstate which substance is known to adsorbe certain peptides,
for instance glutathione. Another important factor which leads
to loss of nitrogen is the splitting of glutamine. We shall return
to this point below.

Table II.

Total N Amino-N

Oryctes =Vn

Total blood 496 284

PW-fraotion found 175 82

Piltrato from PW-fraction 232 164

PW-fraction calculated 264 120

The following conclusions seem to be valid for the blood amino-
acids of both species studied; 1) The dicarbonic acids (the Ba-
fraction) are quantitatively of little importance.

2) The PW-fraction may contain hexonbases and low peptides
(di- and tri-peptides) but hardly higher peptides to any extent.
This is easily seen by comparing total nitrogen and amino-
nitrogen.

3) The monamino acid fraction contains more total nitrogen
than amino nitrogen. This may mean, that this fraction contains
amides or possibly low peptides which have not been precipitated
by the phosphotungstic acid. Of course the excess nitrogen might
originate from substances other than those showing amino acid
reactions. One source of excess nitrogen in Melolontha has, how-
ever, been found already, namely asparagine.

3. The PW-fraction.

In order to see to what extent the PW-fraction consisted of peptides
the follo^ving experiment was made: Half the PW-fraction from Oryctes
~'ln (10 ml) plus 1 ml cone, hydrochloric acid was evaporated in a test
tube on a steam bath until only 2 ml remained. Then the tube was

AMINO ACIDS AND RELATED COMPOUNDS.

sealed and kept at 100° for 18 hours. After that the tube was opened,
the contents evaporated to drjuess and the residue taken up in 10
ml water. After a sample had been dra-svn for amino-N determination
the remaining solution was acidified with a drop of sulphuric acid
and precipitated with phosphotungstie acid in small excess. (Now that
the peptides are split phosphotungstie acid should only precipitate
hexonbases.) The precipitate was freed from inorganic ions. The re-
sulting solutions were used for amino-N determinations. The results
were the following:

Calculated as mg amino-lSr
per g original blood

PW-fraction before hydrolysis 0.82

» » » after » 1.03

Hexonbases in hydrolysate 0.40

Non-Hexonbases in hydrolysate 0,51

It is seen that the amino-N increases 0.21 mg per g blood on
hydrolysis; at the same time 0.51 mg N has appeared which is
no longer precipitated by phosphotungstie acid. This indicates
that the peptides in question are di- or tri-peptides. About half
the amino-N in the PW-fraction comes from hexonbases.

A similar study was made on the PW-fraction of Melolontha
blood (=v«). A suitable part of this fraction was diluted to 5 ml.

0.5 ml from this solution contained 0.655 mg amino-N.

After hydrolysis a corresponding sample
contained 0.692 mg amino-N.

In the remaining part of the hydrolysate histidine + arginine
were precipitated as the Ag-salts and in the filtrate, freed from
Ag+ lysine was precipitated as the phosphotungstate.

Amino-N determinations gave the following results;

Histidine + arginine^ 1.24 mg amino-N

Lysine -fraction 1.39 » »

Filtrate from lysine fraction 0.53 » »

It is seen that peptides contribute to a minor part only to the
amino-N of Melolontha blood. As pointed out in section 2 some
peptide may have been lost by adsorption on Ba-phosphotung-
state.

From the histidine -f arginine fraction histidine and arginine
were fractionally precipitated as Ag-salts according to Stetjdel
(1910). The histidine was lost. The arginine fraction was concen-

^ The Folin method gives much too low values for arginine, whereas lysine
with both amino groups, thus giving per mole about twice the colour- value
of that given by a monamino acid (FotiN 1922).

0-^/60215. Acta pkys. Scandinav. Val.ll.

HAKS H. IJSSIKG.

trakd to one ml at pH 4 — 5 and precipitated -witli 30 mg flaviauic
acid. Tlie orange coloured arginine flavianate -was t\vice recry-
Rtallized from one ml -water containing a trace of fla-vdanic acid.
Eacli time three days in the cold were allowed for the crystalliza-
tion.

Crop: 19.0 mg arginine fla-voanate, corresponding to 6.8 mg arginine.

The lysine fraction, freed from inorganic ions, was concentrated
with a trace of BaCOj to a small volume. It was then freed from
BaCOs by centrifuging and the solution was concentrated to
about 100 //I. Two vol. alcohol, followed by 20 mg picric acid
in hot alcohol were added and the lysine picrate was allowed to
crystallize in the cold.

Crop: 11.2 mg lysine picrate, corresponding to 4.4 mg lysine. It was
left undecided if the lysine fraction contained putrescine, which sub-
stance together with lysine was found bj' Ackermann (1920) in an ex-
tract of whole Jlelolontha lar\'ae.

Prom the above it follows that the greater part of the PW-
fraction from Melolontha consists of the basic amino acids; ar-
ginine and lysine have been isolated and in chapter 7 it will be
shown that histidine is also present in appreciable amounts.

4. The Ba++-fraction.

From table I it is seen that the total K and amino-N of this
fraction is about equal. The fraction therefore probably consists
of monaminodicarbonic acids like aspartic and glutamic acids.
Although the amount of amino-N in this fraction is rather low,
the values are probably still too high. There is good e-vddenco
that some glutamic acid may have been formed from glutamine
during the analyses.

Esjmcially in l\IeloIontha the Ba-fraction was rich in glycogen.
On evaporation a glassy residue was left, which consisted mainly
of this polysaccharid.

6, Amides: Asparagine and glutamine.

As shown in table I the monamino acid fraction contains more
total nitrogen than amino-H. To find out whether this excess N
originated from amides the following hydrolysis was made;

4 ml monamino acid filtrate, (Oryctes ’7«)) corresponding to l.cg
blood and 1 g H.SO, were kept at 100° for 18 hours in a sealed ampoule.

AJIINO ACIDS AND KKLAIND COMPOUNDS. ()7

H 2 S 04 was removed with Ba(OH). and amino-N determinations were
made before and after removal of ammonia

before hydrolysis 0.75 mg N/g blood

after * O.oo » » »

» » and removal of ammonia ... . O.cs » »

It is thus clear that the increase in amino-N on hydrolysis is
due to formation of ammonia. This made it probable, that the
blood contains an amide. This suggestion was strengthened when
it turned out, that the monamino acid fraction gave a precipitate
with mercuric acetate. The experiments were continued on Melo-
lontha blood from which asparagine wms subsequently isolated.
The procedure is described elsewhere (Ussing 1945 a) together
with the analytical data, which should not be recapitulated here.

Only a few points may deserve further mentioning: At first it
seemed impossible, that the substance isolated should be as-
paragine, because it was found that the amino-N (determined
colorimetrically) was only 36.5 % of the total N. Formoltitration
gave, however, the right value, 50 %, and then it turned out that
pure asparagine (kindly supplied from the Carlsberg Laboratory)
gives also too low amino-N values in the colorimetrical method;
the apparent amino-N was found to be 37.3 % of the total N.

A solution of the asparagine isolated wms subjected to carbon
adsorption (Ussing 1945 b). The amino-N decreased 20 % per
carbon treatment, exactly what w'as found for pure asparagine.

, (From the same blood filtrate a small crop of tjnrosine was iso-
lated and identified by the crystal form, low solubility in wmter
and an intense Mllon reaction.)

As mentioned above the presence of an amide in the monamino
acid fraction was first indicated in Oryctes and it was later re-
affirmed (compare Ussing 1945 a). As soon as Oryctes larvae be-
came again available, an attempt was made to isolate as-
paragine from these animals. The blood from 25 larvae, 27 ml,
was used and the monamino acid fraction prepared as usual
(55 ml). The solution was neutralized to lithmus and mercuric
acetate and sodium hydroxide were added by drops until no
more precipitate w'ould form. The precipitate after washing was
suspended in water and decomposed with HjS. The solution was
concentrated practically to dryness, but no asparagine crystallized.
The only crystals formed wmre some fine needles at first suspected
to be tyrosine. The sample was stored for later examination.

During the. preparation of the monamino acid fraction used it

nA>'s n. uspiNO.

was noticed, however, that the addition of hot Ba(OH); solution
to remove phosphotungstic acid produced a strong smell of am-
monia although the blood does not contain this substance to a
measurable extent. This means that a substance is present, which
splits off ammonia in alkaline solution even at temperatures
appreciably below 100°. It could hardly be asparagine because
this substance is relatively stable and it was therefore suggested
that the substance in question was glutamine. Whether gluta-
mine or another instable substance it was evident that the
method hitherto used was likely to destroy a considerable part of
the substance. Treatment with hot alkali should be omitted and
it was therefore decided in the next experiment to precipitate
the supposed glutamine before the PW-precipitation. (Another
possibility was to remove excess phosphotungstic acid by ex-
traction with n-amylalcohol and ether; but it has been impossible
to get n-amylalcohol during the war.) The most simple way would
have been to precipitate amides with mercuric nitrate after re-
moval of interfering substances with Pb-acetate (compare for in-
stance Vickery et al. 1935 a); but in this way a considerable
amount of nitrate ion would be introduced which could not be
removed again, and the filtrate from amide precipitation had to
be reserved for other analyses (see below). Tests with asparagine
and urea (glutamine was not accessible) showed that precipitation
with Hg-acetate was not quantitative in dilute solutions as is
the precipitation with Hg-nitrate. The precipitation of asparagine
with Hg-acetate was, however, much improved when 3 volumes
alcohol were added, Avhile on the other hand this treatment did
not precipitate most monaminomonocarbonic acids.

Tlie procedure adopted was therefore as follows: Blood-filtrate was
prepared as usual from 26.9 g’s of Oryctcs-blood (Oryctes "j^). After
concentration in vacuo to 50 ml, lead acetate was added until no more
precipitate was produced. The solution which was first acid to Kongo-
paper became neutral (due to precipitation of the sulphate and phos-
phate ions). After removal of the precipitate, mercuric acetate was
added until no more precipitate would form. Then 3 volumes alcohol
were added and the precipitate allowed to settle in the cold overnight.
The precipitate was washed with 80 % alcohol, suspended in water
and decomposed with ILS. H.S was removed by concentration in
vacuo. Some contaminating substance (histidine ?) was precipitated
with a little phosphotungstic acid after addition of some drops of sul-
phuric acid. Jlost of the H;S 04 was removed with Ba(OH) 2 . The last
trace of H-SO^ and the e.xcess phosphotungstic acid was removed with
lead acetate. Lead was precipitated with HoS. The solution was con-

AMIXO ACIDS AND REDATED COMPOUNDS.

centrated in vacuo to 2 ml. It was then transferred to a small porcelain
dish and a little ammonia was added to neutralize the acetic acid. By
gentle heating on a waterbath the solution was concentrated in a
strong air current to speed uj) the evaporation without undue heating
(glutamine decomposes rapidly even at neutral reaction, when the
temperature is above CO''). An oily remancncc was left: about .C vol-
umes of alcohol were added and after vigorous stirring for several
minutes the oily rcmanence would cry.stnl!ize in very small needles.
Tyrosine was only present in traces (Arnow’s modification of the Jlillon
reaction (1937)).

According to Vickery cl al. (1935 b) glutamine splits off all
or nearly all its amino-N when it is heated to 100“' for 2 hours at
pH 0.5. Asparagine gives no ammonia under these conditions and
even urea and allantoine which are relatively unstable are only
partly split in two hours, so that heating for four hours will give
much more ammonia than heating for two hours. These authors
suggest that these facts may be used to differentiate between the
amides mentioned.

The following c.vpcriment was made: IS.o."? mg of the powder, sup-
posed to contain glutamine was dissolved in 7 ml water. One ml samples
were mi.\ed with 1 ml each of phosphate buffer, pH C.r!, in test tubes
which were sealed off and heated in boiling water for two and four
hours. Then the tubes were placed in an ice-bath and the water con-
densed in the upper end which may contain ammonia driven down
by gentle heating with a free flame. The ampoules were opened and the
contents transferred to the distillation flask of a distillation apparatus
of the Parnas type together with the washings, 2 ml. 2 ml 1 % NajCOj

were added and the ammonia distilled over in 1 ml ILSOj.

Backtitration with ^ NaOH. Indicator: Mi.vture of mothylrcd and
meth}’lenc blue. Results arc .shown in table III.

Table III.

^ 100

1 ml sample -f- 1 ml buffer, not heated

* * ' 4- » » » , heated for 2 hours

• * * -p » » » , heated for 4 hours

In two hours thus the equivalent in NHa of 72.3 /d NaOH was

found, whereas in four hours the titnation gave 79.2 /d. This

<d NaOH used

.... 187

I 93.8
■••• \9G.2
f2G.i
•••• \24.0
f 19.f.

•••• tlC.G

UANS H. DSSING.

means that more than 90 % of the ammonia was released witliin
2 hours.

In order to purify the supposed glutamine, the remaining sample
was boiled with an excess of freshly prepared Cu(OH) 2 , filtered
hot and concentrated to a small volume. The next day the syrup
had crystallized. The crystals were digested twice with 0.3 ml
water and dried: 35.4 mg

15.0 mg copper salt was dissolved in water and decomposed
with HjS. After removal of H-S the solution was made up to
10 ml and the glutamine content was determined as above with
the exception that 1 ml 5 % borax in 0. 5 n NaOH was used in-
stead of NajCOj for alkalizing. The results were (table IV):

Table lY.

1 ml ^ H.SO,

1 ml sample + buffer, not heated / . . .

» t » , heated for 2 hours . .

.» » » 4- » , heated for 4 hours . .

In two and four hours respectively the NHs-equivalent in NaOH
of 67.9 /d and 72.2 /d were released. This means that about 94%
decomposition takes place within 2 hours. It is seen that the sub-
stance contains still some ammonia, the equivalent of 20 /d NaOH
per ml solution, and correspondingly the content of amide-N is
somewhat low, 6.4 % of the copper salt instead of 6.7 % as cal-
■culated for glutaminecopper. Further recrystallization was aban-
doned on account of the small amount of substance left.

The substance isolated by Hg-acetate precipitation from Oryc-
tes blood, monamino acid fraction (see above), was now dissolved
in water and aliquotes heated at pH 6.5. It proved that this
sample also contained glutamine: the ammonia set free in 2 hours
was 98 % of that formed in 4 hours.

Although glutamine was not isolated in the pure state, its pres-
ence in Oryctes blood must be regarded as demonstrated. Its
identification rests on the following facts: The blood contains an
amide, which is not precipitated by phosphotungstic acid, lead
acetate or baryumhydroxide and alcohol. It forms a copper salt,
which is but little soluble in water (difference from urea, allan-
toine a. o.) and it splits off nearly all amide-N within 2 hours

«1 NaOH used

.... 187

1157.0
•••• 115G.7
; 90.O
• • ■ • \ 88.5
f 85.4
••••1 86.5

AMINO ACIDS AND REDATED COMPOUNDS.

ivlien heated to 100° at pH 6.5 (difference from asparagine, urea
and allantoinc). It is insoluble in alcohol and crystallizes in fine
needles. Even the tendency to form supersaturated solutions is
characteristic (compare Vickery ct al, (1935 a)).

The next question was if glutamine is also found in the blood
of Jlclolontha.

Blood was taken from 25 larvae (--/j) and the protein- and trichloro-
acetic acid free extract was prepared as usual. The solution was acidi-
fied vrith sulfuric acid and bases and peptides precipitated with phos-
photungstic acid. SO* and phosphotungstic acid were removed with
lead acetate and load with IIjS. Acetic acid was removed after con-
centration by extraction with ether, 4 ml solution, corresponding to
l.n g blood were mixed witii 4 mi buffer (pH. C..'>). Using a glass elec-
trode pH was adjusted e.xactly to 6.5 with ~ NaOH. 2 ml samples

of this solution each corresponding to 0.241 g blood were heated as
above for 2 hours. Ammonia was formed corresponding to O.iss mg H
per blood, or to 19G mg % glutamine.

The rest of the blood filtrate was freed from a trace of phospho-
tungstic acid with Bn(OII)s and Bn++ removed ndth H,SO*. The re-
sulting solution was adjusted to pH 6.5. G two ml’s samples were
mixed with 1 ml buffer (pll 6.5) each and two 2 ml samples were mixed
with 1 ml 10 %-il 5 S 04 . The further treatment and the results arc seen
from table V: each sample corresponds to 0.4 1 g blood.

Table V.

1 jog H,.SOj

pH 6.5; not licntcU

pH 6.5; 2 hours’ heating

pH 6.5; 4 liours’ heating

Strongly acid; 3 hours' healing .....

Blank on 1 ml 10 % sulphuric acid

The samples containing 10 % HjSO* were made alkaline with
a mixture of the equivalent amount of NaOH and the usual
amount of borax-buffer. The ammonia set free in 2 hours at pH
6.5 is 97 % of that formed in 4 hours. What is of special interest
in this case is, that when correction is made for the small NHj
content of the sulphuric acid, practically the same amount of
ammonia is set free at pH 6.5 and under the much more strenu-
ous hydrolysis with sulphuric acid. But this means that these

/<! NaOH u.sed

.... 187

1168.5

V oo.c*

I 78.0
1 79.2
I 68.1
\ 66.0
180

HAKS H. TJESIKa.

larvae contain glutamine as the only amide. No or practically no
asparagine is present in this case.

A similar experiment on Oryctes blood filtrate gave an amide-N
content of 0.142 mg/g blood, corresponding to 150 mg% glut-
amine.

6. The behaviour of blood amino-N on carbon-adsorption.

In a previous pajjer (Ussikg 1945 b) the decrease in amino-N
on repaeted adsorptions on carbon under standard conditions vas
used to characterize certain amino acids. This method may be
used to find out if an amino acid mixture contains to any extent
glycine, alanine or other little adsorbable amino acids.

The procedure is described in detail in the paper mentioned above.
Suffice it to say that in each treatment 30 mg carbon per ml solution
vas used so that the adsorption was onl}' a function of the nature
of the amino acids in question and of the concentration. It was further
found that in the very dilute solutions used (1 — 5 mg % amino-N) the
adsorption was approximately proportional to the concentration, or
in other words the relative adsorption was independent of the con-
centration.

Table VI.

% adsorption per
carbon treatment

Glycine 1.5

Serine 6

Alanine 6

Asparagine 20

Glutamic acid 20

Ly.sine 24

Valine 27

Leucine 60

Histidine 84

(Glucose 70)

Table YI shows the approximate deereaso per carbon treatment
of some amino acids; glucose in similar concentration is also in-
cluded for comparison, pH in all cases 7.

DIany such adsorption analyses were made on blood extracts
from Oryctes and DIelolontha, on fractions as well as on unfrac-
tioned blood filtrates. In all cases it was found that the mon-
amino acid fraction contained a considerable proportion of little
absorbable amino-N. Examples are shown in fig. I. Apart from
adsorption curves for monamino acid fractions from Oryctes and
Jlclolontha, the figure shows the curve for unfractionated blood

amino acids and hedatkd compounds.

I'ig. 1.

filtrate from imagines of tlic grnsliopper Locustn viridi.s.sinm. In-
cidently the curves linppen to liavc the endpoint in common and
this nlloM's an easy comparison between them. All adsorptions arc
performed on very dilute .solutions, but the resultsS arc recalcu-
lated to the original concentration in the blood. It is seen that
after 5 carbon treatments which remove most amino acids, the
curves for the three species are nearly identical. The decrease per
carbon treatment is between 4 and 5 %.

The presence in the blood of Oryctos, Melolontha and Locusta
of one or more of the little adsorbablc amino acids in a fairly high
concentration is thus ascertained.

The decrease, 4 — 6 % per carbon treatment shows, however,
that the amino-N left after 5 treatments does not come from
pure glycine (decrease 1.5 %). The substances, which may come

HANS H. USSns'G.

into consideration, are apparently glycine, serine and alanine,
moreover are proline and hydroxyproline known to be very little
adsorbed by carbon (Tiselius 1941). Attempts were therefore
made to identify one or more of these amino acids.

Serine. According to Rapoport (1937) this amino acid may be
estimated by being transferred by deamination to glyceric acid,
which gives a specific reaction with naphtoresorcine. Dicarbonic
acids and hexonbases should be removed. Sugars too interfere.

The blood filtrates (monamino acid fraction) gave no Fehling
reaction; but with a-napthole and sulphuric acid they gave a
\'iolet colour. It proved, however, that three treatments with
carbon (see above) removed the sugar as estimated by the
a-naphthole reaction, whereas serine, if present, should only be
reduced some 14 %.

Several determinations were made both on Oryctes blood and
on the blood of the larvae of the beetle Rhagium mordax; but in
no case the blue colour characteristic for serine was obtained. A
small content of serine would, however, escape detection because
a brown colour always occurred, when insect blood was examined.
A similar colour was also produced by some amino acids as for
instance glycine and alanine.

It thus seems that serine, if present at all, is quantitatively un-
important in Oryctcs blood.

Alanine. Forth, Scholl and Hermann (1932) determined
alanine in protein hydrolysates by converting it into lactic acid
by deamination. The lactic acid was determined according to
Friedmann and Kendall (1929). Dicarbonic acids and the bases,
which may also give rise to some lactic acid should be removed
before the deamination. Since this method was worked out
threonine has been recognized as a constituent of many proteins,
and this amino acid will probably also form lactic acid under the
above conditions.

As no Oryctes larvae were available at that time, determinations
were only made on blood from Melolontha. As an example the
following experiment may be cited:

A monamino acid fraction was prepared as usual. 25 ml of this
fraction, corresponding to 12 g’s of blood were treated twice with
carbon (see above) which removes impurities without adsorbing much
alanine. From the resulting solution 10 ml were used for direct lactic
acid determination and 10 ml deaminized and then used for lactic
acid determination, thus giving the sum of preformed lactic acid and
lactic acid formed from alanine (and threonine)

AMIKO ACIDS AND REDATED COMPOUNDS.

10 ml not deaminized used .■ l.os ml

20 Jill 9 2 . so ml 1 ) ft

lactic acid from alanine thus used 1.15 ml » »

This corresponds to about O.oa mg alanine or 0.2 mg alanine per g
blood.

Even this low value is, however, too high because, as was later
found the monamino acid fraction contains glutamine and some-
times asparagine, both of which will give rise to lactic acid on
deamination and oxydation. It thus seems that alanine (and
threonine) are only present in Melolontha blood in quite small
amounts, if at all.

Hydroxyprolme. Glycine, proline and hydroxyproline form char-
acteristic picrates. Especially the picrates of glycine and proline
are Uttle soluble and should be suitable for the isolation of those
amino acids. Therefore the filtrate from the glutamine precipitation
on Oryctes blood (Oryctes see above) was used for the attempt
to identify as the picrates one or more of the amino acids mentioned.

The alcoholic filtrate was concentrated in vacuo to remove al-
cohol. Hg‘‘‘+ was removed with H.S and HjS by distillation. Bases
were precipitated with phosphotungstic acid after addition of
some drops of sulfuric acid. The filtrate was freed from phospho-
tungstic and sulfuric acid wdth an excess of Ba(OH)2 and the
Ba++ was exactly precipitated with H-SOj. Acetic acid which
originated from the Pb- and Hg-acetates was extracted with ether.
After removal of ether by vacuum distillation the solution was
made up to 100 ml and shaken with 6 g carbon (Schering’s Garbo
ossium pro analysi). After filtration the carbon treatment was re-
peated with 5 g’s of carbon. The filtrate from the second treat-
ment noAv neutral to lithmus was concentrated to dryness in a
porcelain dish. The residue was extracted with a minimum of hot
water and the resulting solution (about 200 //I) was mixed with
5 volumes of alcohol. A precipitate formed which proved to con-
sist of inorganic sulfates, whereas all amino-N remained in the
alcohohe solution. This solution was concentrated to a syrup
(50 /<1), but only a few cubic crystalls of alkali chloride appeared.

The syrup was dissolved in two ml water and from this solution
aliquotes were taken for amino-N determinations according to
PoLtN (modified, see Ussing, 1. c.) and van Slykb.

Polin; 4.80 mg N in whole sample; van Slyke; 1.28 mg N in
whole sample.

HANS H. USSING.

Tlie difference 3.52 mg N must originate from amino acids
which are determined with the colorimetric method and not ■with
the van Slyke method. From the known amino acids this is only
the case with proline and hydroxyproline and it is thus strongly
indicated that one of these amino acids — or both — are jiresent
in Oryctes blood in considerable amounts.

Grassmakn (1935) has described a characteristic reaction for
these two amino acids. When they are heated in watery solution
with a little isatin a blue colour is produced. The coloured sub-
stance is eagerly adsorbed on fibers of cotton or acetate-cellulose.

A drop of the solution which was being examined was boiled
with a little isatin for two minutes with 1 ml phosphate buffer
(pH 7) and the solution soon turned green. A tuft of cotton wool
was placed in the solution, boiled and washed with hot water; it
had turned intensely blue.

1.4 ml still remained of the “little adsorbable amino acid frac-
tion” (Oryctes -j^). It was concentrated to 0.5 ml in a small glass
tube, using an air current to speed up evaporation. 40 mg picric
acid were added under stirring at 100°. On cooling crystalhzation
begun and the whole sample became a mass of long needle-hke
crystalls. Under the microscope they appeared exactly like hyd-
roxyproline-picrate with the characteristic tendency to form
“brushes” (see Crosby and Kirk 1935). The substance was rather
soluble in water in contrast to proline-picrate. The crystals were
sucked off on a micro glass filter, dried and washed ■ndth ether;
crop: 30 mg, 26.8 mg picrate ■were dissolved in 8 ml water and
after addition of 3 drops of 10 % H2SO4 the picric acid was ex-
tracted Avith ether. After removal of H2SO4 and ether aliquotes
of the solution were oxydized with hypochlorit and subjected to
a vapour distillation all according to AValdschmidt-Leitz et al.
(1934). Under these conditions from all known amino acids only
hydroxyproline gives pyrrole to be shown by a violet colour
which develops when dimethylaminobenzaldehyde and hydro-
chloric acid is added to the distillate. The reaction Avas positive
for the picrate-fraction of Oryctes blood and so the presence of
hydrox}q)roline Avas finally demonstrated.

Untreated blood filtrate from Oryctes AA-as also tested for hy-
droxyproline according to Wahdschmidt-Leitz and Avith strongly
positiA'c result but due to lack of a hydroxyproline standard the
test could not be used for quantitative determinations.

AVhether proline is also present has yet to be found out; but

AMINO ACIDS AND RELATED COMPOUNDS.

the liigh solubility of the picratc and the crystal form speaks
against the assumption that more than a trace of this amino acid
is present. In another experiment in which picrates were isolated
from Oryctes blood in essentially the same way as that just
described, the picrates were recrystallized from hot water, with
the result, however, that only the picrate of an inorganic ion,
presumably potassium was left. Thus no insoluble or little soluble
amino acid picrates like those from proline or glycine were found.

The fact that the Oryctes blood filtrate even after carbon ad-
sorption gives some amino-N in the van Slyke apparatus indicates
that beyond hydroprolinc other little adsorbable amino acids arc
present in minor amounts. .Due to lack of material similar analyses
have not been made on Mclolonthn blood.

7. The concentration of certain amino acids in blood and
intestinal contents of Melolontha.

Blood extract from 17 hiclolonthn larvae (■l..'> g blood) was pre-
pared in the usual wa}'. The blood ])rotein was filtered off on a weighed
ash-free filter, washed with water and dried to constant weight at
105°: O.ino g or S.ar, %.

Contents from the mid-gut was collected in the following way: A
sagittal incision was made in the skin in the midlinc of the back of the
larvae. As a rule the midgut would then protrude in a hernialike fashion.
Blood adhering to the gut was removed with filter pn])cr and the con-
tents of the gut could be coliected from an incision in the gut wall
in to a weighed vc-ssel, containing 2 ml of 20 % trichloroacetic acid.

2.24 g gut-content was taken and then the solution was made up
to 22.1^ ml and filtered. The insoluble rc.eiduc on the filter was washed
and dried: 0.2s:) g or 12.c.'i %. The filtrate freed from trichloroacetic
acid in the usual way was quite dark from some huminlike substance.

The colour was, however, practically removed by precipitation with
baryumhydroxidc and the remaining colour was adsorbed on the
BaSO^, when the Ba+ + was removed with ILSOj.

For the sake of comparison the blood filtrate was likewise treated
with Ba(OH)2 and ‘then freed from ]3a++ again.

On the filtrates thus prepared analyses were made for total
amino-N, histidine (according to Rackur (1940)), tryptophane
(according to Winker (1934)). tyrosine (according to Arnow
h c.) and leucine -f valine (Ussing, 1943). The results of these
analyses are presented in table VII, column I and II. In column
III and IV the results arc calculated on the basis that the dry
substance, insoluble in trichloroacetic acid is not acting as solvent
for the amino acids. Nor the blood an estimate is given for leucine

HAKS H. USSING.
Table TH.

Xiilustiiuul

Corrected for dry substance

;

Blood
mg %

contents,

mg %

Blood
mg %

Intestinal

contents,

mg %

Intestinal

contents

Blood

Amino-N

Histidine

Tryptoplian

Tyrosine

Leucine + valine (ns

leucine)

Leucine

Valine

Hicarbonio acid N . . .

342

300

12.5

lOG

226

131

31.2

67.5

354

77.3

c. 100

311

c. 112

110

234

—

32.4

0.22
C. 0.30
0

0.38

0.33

and valine separately; the procedure used is described in the next
section.

IFrom the table it is seen that all amino acids examined are pres-
ent in higher concentration in the blood than in the gut. In an-
other case ("‘ho) amino-N determinations were also made on in-
testinal contents. The concentration ivas 48.8 mg% against 453
mg% in the blood. We shall return to this point later.

8. Differential estimation of leucine and valine.

In their original method Fromageot and Heitz (1939) used
the different amounts of acetone formed on short and prolonged
chromic acid oxydation from deaminized leucine and valine for
the differential estimation of the two amino acids. Block (1940)
used oxidation with two different oxydants (cliromic acid and
permanganate) for the same purpose.

When very small amounts of the two amino acids are to be
determined, these methods present serious difficulties, and it was
therefore attempted to modify the method worked out by the
author (UssiKG 1943) for the determination of the sum of leucine
and valine so as to permit differential determinations.

As shown in table VI valine is much less adsorbed by carbon
than is leucine, and this fact is made use of in the estimation of
the two amino acids.

The best differentiation proved to result when GO mg carbon
was used per ml solution.

The estimation is made in the following way: From the solution
to be examined a sample containing 0.5 — 1 mg leucine + valine

AMINO ACIDS. AND BELATED COMPOUNDS.

is taken for the direct determination of tke sum of these amino
acids. Another sample containing 2 — 4 mg leucine + valine is
neutralized to lithmus and diluted to 10 ml. 600 mg carbon is
added and the mixture shaken thorouglily. The carbon is filtered
off and the filtrate is measured •with an accuracy of 0.05 ml,
transferred to a test tube, deaminized, concentrated and oxidized
in sealed ampoules upon -which the distillation of acetone and the
colour development is performed in Conway units as usual.

Tor the calculation the folio-wing empirically found facts have
been used: 1) Equimolar amounts of leucine and valine give equal
amounts of acetone on direct determination. 2) Leucine decreases
to 13.5 % on carbon treatment whereas 3) valine decreases to
41.5 %.

Fig. 2.

Tig. II shows the influence of adsorption on known leucine-
valine mixtures. The ordinates (in arbitrary units) are corrected
for blank and calculated under the assumption that the molar
equivalent of 0.5 mg leucine is analyzed each time. Without ad-
sorption 0.5 mg lucine or the equivalent amount of vahno would
give 320 arbitrary photometer units.

HANS H. USSING.

The error on the valine or respectively the leucine values may
sometimes exceed 10 % but still this method allows an approximate
determination of the amino acids in question in extremely small
samples.

9. Variations in total amino-N in the blood filtrates.

Considerable variations were found in the non-protein amino-N
of the blood filtrates. Table ^nil shows examples of these varia-
tions together with remarks about the conditions under which
the samples were taken.

Table YIII.

Animals directly from habitat (Oryctcs ”/}) 31G

» » » » ( » 219

Animals kept at room tp. (20 — 22°) for 11 days (Oryctes =’/„) 284

Animals directly from habitat (Slelolontha “Ao) 375

i> » * » ( » =’/io) 453

Animals kept at room tp. for 10 days (Melolontba '*/io) 342

» » * f » 9 » » ( » •■'/,) 280

Discus.sion.

In the two species studied, Oryctes and Melolontha, the non-
protein amino-lsT of the blood has been found to originate from
three main groups of substances: Low peptides (di- and tripepti-
des), amides (glutamine and sometimes asparagine) and real
amino acids. Nearly all nitrogen can be accounted for as belonging
to the groups of substances mentioned.

In Slelolontha blood the presence of the following substances
has been established: The bases lysine, arginine and histidine, the
monoaniino acids tjTosine, leucine, valine and tryptophane and
the amides glutamine and asparagine. From the shape of the car-
bon adsorption curve (see fig. I) the presence of a considerable
amount of hydroxyproline is indicated. ^Moreover a small amount
of peptide and possibly some dicarbonic amino acid is found. Of
these substances lysine, arginine and leucine were isolated by
Acki-:rmann (1920, 1921) in extracts from whole Melolontha lar-
vae, but as the high amino-N content of insect blood was not
known at that time, the amino acids mentioned were apparently
assumed to be tissue extractives. Ackebjiann did not isolate
tyrosme. Probably it had been destroyed by enzymes in the

AMINO ACIDS AND KELATED COMPOUNDS. 81

tissue mince. Perhaps the p-hydroxyphenylethylamine isolated
by Ackeemann had been formed in this way.

In Oryctes blood conditions are quite similar as far as the anal-
yses go. A quantitative difference is found in the monoamino
acid fraction, which is in this species apparently dominated by
hydroxyproline and glutamine. The moderate decrease on carbon
adsorption (see fig. I) shows that the concentration of the strongly
adsorbed amino acids, for instance leucine and tyrosine must be
low and this fits in with the fact that no tyrosine could be isolated
from the Hg-precipitates from Oryctes in contrast to that from
Melolontha.

It cannot be ruled out, however, that seasonal changes may
have something to do vith the differences found. In this connec-
tion the curious fact that asparagine was found in Melolontha in
the autumn but not in the spring may be mentioned. The dif-
ference may be due to changes in metabolism; but another pos-
sibihty is that asparagine may be taken up by the animals from
the plants on which they feed and stored for some time in the
blood before it is split or excreted.

The glutamine, which is found in both species, may of course
originate from plant-food too, but its presence might just as well
have some connection with the syntheses of uric acid. As shown
by Okstrom, Orstrom and Krebs (1939) the uric acid formation
in liver slices from birds is speeded up on glutamine addition.

The question put in the introduction: are insects able to trans-
fer amino acids from a low concentration in the gut to a higher
concentration in the blood, can now be answered in the affirm-
ative with some certainty. Not only the total amino-N, but also
the individual amino acids analyzed for are much more con-
centrated in the blood than in the intestinal contents (see table
VII)j even when correction is made for the higher content of
dry matter in the gut. As most of the osmotic pressure in insect
blood is due to amino acids it seems likely that the blood and the
contents of the gut are very far from isotonic. It is a type of
absorption which reminds of the reabsorption of substances in
the tubules of the kidney of vertebrates. It is possible that the
conditions found are peculiar for the very curious digestive system
of the lamellicorn beetles. Studies on other types are needed.

The total concentration of amino-N in the blood is subject to
considerable fluctuation in both species studied (see table VIII).
Nothing definite can be said about the reasons for these fluctua-

6—460215. Acta phys. Scandinav. Vol. 11.

HANS H. OSSING.

tions, but as a working hypothesis it is supposed, that temperature
has something to do wth the phenomenon. Taking Oryctes first,
it is seen that the concentration is lowest when the outer tempera-
ture is highest. The reason why the animals used -=/ii had been
kept at room temperature was that due to liigh viscosity practic-
ally no blood could be obtained 10 days before when the animals
had been kept at 10° for some time.

In the case of ]\Ielolontha special interest should be paid to
the analyses for the dates =’/io and ■'/lo- These two days and again
the animals were collected on the same locality. In the course
of that week the weather became cold and the ®'/io only 2 animals
were found. The others had left the upper strata of the earth and
dug down to the wintering level. In this week the amino-N in-
creased from 375 mg % to 453 mg % and the total non-protein
N from 447 to G43 mg %. An excretion of water (and possibly
some amino-hl) would be the simplest explanation on this phenom-
enon. That insects concentrate their blood under diapause is
well established (compare Mell-ANBY, 1938), but possibly a more
general dependence of outer temperature exists.

Although a high amino-N content is common to all insects
studied there are apparently considerable differences as to the
relative concentration of the individual amino compounds be-
tween the different insect species. Flokkin and Duchateau (1. c.)
were unable to find for instance arginine and tryptophane in
adult Dytiscus, whereas both amino acids were found in Melo-
lontha larvae. Perhaps the reason is that the haemolymph of the
larvae is acting as a deposit for substances, which are used for
protein synthesis during the metamorphosis (compare Heller,
1932). Arginine moreover may leave the blood to be stored in the
muscles as phosphoarginine.

The adsorption curve for unfractioned blood from Locusta viri-
dissima (fig. I) shows clearly that no single amino compound
dominates the picture in adult insects. At least two and possibly
more substances are of importance. The slope of the curve be-
tween the 5’ and the 10’ carbon treatment justifies the search for
hydroxyproline in this species too. hluch more work has to be
done, however, before a clear picture of the amino acid com-
position of insect blood has been obtained.

The author wishes to express his gratitude to Miss Gierloff
who has performed much of the analytical work.

AMINO ACIDS AND REDATED COMPOUNDS.

Summary.

1. The haemolymph of the larvae of the beetle Melolontha vul-
garis was found to contain between 280 and 450 mg % non-
protein amino nitrogen; in haemolymph from the related Oryctes
nasicornis between 219 and 316 mg % was found.

2. This amino nitrogen originates from three main sources,
namely a) real amino acids, b) low peptides (di- and tripeptides)
and c) amides (glutamine and asparagine).

3. In Melolontha blood the presence of the following substan-
ces was established; lysine, arginine, histidine, tyrosin, leucine,
valine, tryptophane and the amides asparagine (in some cases)
and glutamine; the presence of hydroxyproline in an appreciable
amount was indicated, liloreover a small amount of peptides and
some dicarbonic amino acid or acids was present.

4. In Oryctes blood hydroxyproline was identified; this sub-
stance and the basic amino acids play the most important part
among the real amino acids. Low peptides are present in consider-
able amounts. Glutamine seems always to be present, whereas
asparagine has not been met with.

5. The non-protein amino-nitrogen in the content of the mid-
gut of Melolontha has been examined. It was found that the con-
centration of all amino acids analyzed for was much lower in the
gut than in the haemolymph.

6. The variations in the amino-N concentration of the haemo-
lymph is briefly discussed.

Literature.

Ackermann, D., Z. Biol. 1920. 71. 193.

Ibidem 1921. 73. 319.

Arnow, L. E., J. Biol. Chem. 1937. 118. 531.

Block, R. J., Proc. Soc. exp. Biol., N. Y. 1940. 45. 289.

Crosby, B. L., and P. L. Kirk, Microchemie 1935. 18. 137.

Duval, M., P. Portier, and A. Cokrtois, C. R. Acad. Sci. Paris
1928. 186. 652.

Elorkin, M., et G. Duchateau, Bull. Acad. r. Belg., 01. Sci., V. s.
1942. 28. 373.

Eolin, 0., J. Biol. Chem. 1922. 51. 377.

Eriedmann, T. E., and A. J. Kendall, Ibidem 1929. 82. 23.
Eromageot, C., and P. Heitz, Enzymologia 1939. 6. 258.

Eurth, P., R. Scholl, and H. Hermann, Biochem. Z. 1932. 251. 404.

HANS ir. USSING.

Grassmann, W., and K. v. Arnim, Ann. Chem. 1935. 519. 192.
Heller, J., and A. Moklowska, Biochem. Z. 1930. 219. 473.
Heller, J., Ibidem 1932. 255. 205.

Mellanby, K., Parasitol. 1930. 30. 392.

Hacker, E., Biocbem. J. 1940. 34. 89.

Karorort, S., Biochem. Z. 1937. 289. 406.

Steudel, H., Handb. biochem. Arb. Meth. 1910. II. 498.

Tiselius, a.. Ark. Kem. Mineral. Geol. 1941. 15 B. No. G.

UssiNG, H. H., Acta physiol, scand. 1943. 6. 222.

— , Nature 1945 a. 155. 481.

— , Acta physiol, scand. 1945 b. 9. 193.

Waldschmidt-Leitz, E., and S. Akabori, Hoppe-Sevl. Z. 1934. 224.
187.

Vickery, H. B., G. "W. Pocher and H. E. Clark, J. Biol. Chem. 1935 a.
109. 39.

Vickery, H. B., G. W. Pucker, H. E. Clark, A. C. Chibnall and R.

G. WiSTALL, Biochem. J. 1935 b. 29. 2710.

Winker, S., Hoppe-Seyl. Z. 1934. 228. 50.

Orstrom, a., M. Orstrom and H. A. Ivrebs, Biochem. J. 1939. 33.
990.

From the Physiology Institute at Helsinki University and the Psycho-
Physiological Institute, Defence Forces, Helsinki.

Tlie Influence of Anoxia on tlie Gastric
HCl-secretion.

K . HARTIALA and M. KARVONEN.

Received 23 October 1945.

The gastric HCl-secretion varies sensitively owing to many dif-
ferent reasons. The classical experiments of Pavlov, concerning
conditional reflexes, give light to the part that the nervous system
and physical actions play as regulators of the secretion, but the
humoral agents regulate it also; thus gastrin, the hormone secreted
in the pyloric part of the stomach, acts through blood as an irri-
tant of the HCl-secretion. Some other alterations in the composi-
tion of blood have also influence upon the secretory function of the
stomach. It has also been ascertained, that the HCl-secretion
decreases, when the organism becomes exposed to oxygen defi-
ciency and alterations caused by it. Bayexjx (1910) observed a
noticeable decrease in the quantity of secretion and a smaller
one in the total acidity on a dog, when given a meal of meat in the
height of 4,360 m. Similar results have, using dogs, later onDELRUE
(1934), Sleeth and van Liere (1936) as well as Pickett and van
Liere (1941) attained. Warren (1937) and Hartmann, Hepp
and Left (1941) have also observed, that the HCl-secretion pro-
duced by an alcohol test-meal given to the members of some
mountain expeditions decreased already in the height of 4,000 m.
Hartmann, Hepp and Left observed besides, that the duration
of the HCl-secretion decreased, when the height increased. —
Hellbbrandt, Brogeon and Hoopes (1935) on the other hand,
when investigating the influence of a short duration of anoxia,
came to a so far differing result, that “. . . acute anoxemia of the
pre-coma type has relatively little inhibiting effect upon the secre-

K. JIARTIALA AND M. KARVONEN.

tion of hydrochloric acid by the normal human stomach . .
even though an evident influence is noticeable in their experi-
ments too.

Both Warren* and Hartman'N*, Hepp and Lupt consider the
alkalosis caused by anoxia as the reason of the decrease of the
HCl-secretion; in this case the HCl-secretion implies an extra
transfer of acid ions from blood, already alkaline, and the increas-
ing of alkalosis. It is likewise known, that a hyperventilation,
voluntarily brought about, causes a great decrease in the HCl-
secretion of the stomach. (Delhougxe, 1927; Browxe and
ViNEBERG, 1932).

Delrue in his experiments, which he performed in the height
of 3,64r0 m, using histamine as the irritant of HCl-secretion, came
to such results, that, when the dog becomes acclimatized to anoxia,
the HCl-secretion, at first decreasing, in 2 — 3 hours regains its
earlier level; the return of the HCl-secretion is accelerated by
intense physical work. He does not think, however, that the re-
duction of HCl-secretion is caused by the immediate rise of the
pH of the blood, because NaHCO^, given intravenously, does not
correspondingly decrease the HCl-secretion in spite of the rise
of the pH of blood.

While Jalavisto and the authors were investigating the HCl-
secretion on decompensated heart patients,^ there arose the ques-
tion, whether the oxygen deficiency as such — irrespective of
simultaneous variations in the ion-equilibrium — has a decreasing
effect upon the HCl-secretion. Van Liere (1911) presents as his
opinion, that “. . . the gastro-intestinal tract has a low energy
requirement for secretion, since it is capable of withstanding
relatively severe grades of anoxia, before it is significantly affec-
ted . ”, but he remarks also, that the question has been as yet
insufficiently investigated.

The main purpose of this investigation was to throw experi-
mental light on the influence of anoxia on man, and the plan was
following: the subjects were to be brought in a condition of oxygen
deficiency — e.g. by lowering the barometric pressure — but the
accompanying production of alkalosis was to be prevented. There
are three different possibilities;

(1) mechanical prevention of hyperventilation;

(2) increasing the COj-content of the inspired air;

(3) giving of “fixed” acids to the subject.

' As yet unpublished.

INFLUENCE OF ANOXIA.

The first of the above racutioncd possibilities is of course out
of question, as the experiments arc made on man; the increasing
of CO. as such increases HCl-sccrction (Bakaltschuk, 1928;
Mosonyi, Gunther and Petr.anyi, 1935) and causes a maximal
h}'perventilation, ivliich partly compensates even a serious oxygen
deficiency of inspired air. Giving of "fixed” acids — and salts
acting as acids in organism — on the other hand causes liyper-
ventilation in a relatively smaller degree (Dougla.s, Greene and
Kergin, 1933) and does not as such influence the HCl-secretion,
as Maceagan (193‘1) with NH,C1 and Schifflers (193G) with
H3PO4 have experimentally proved.

Jlcthod.

The experiments were performed using throe healtliy medical students
in the age of 19 — 27 years; IC. H., T. L. and >1. K. The test-meals
were done in the morning about 2 hours after awakening, when the
stomach was empty. After the stomach-tube had been swallowed, the
stomach was emptied 4 times at the intervals of 10 minutes, after which
300 cem 5% alcohol, dyed with methylene blue, was poured in ns irri-
tant. After that si.x samidcs of gastric juice were taken likewise every
10 minutes; ns long as the blue colour was observable, 10 cem was taken
at a time, and when the samples had changed colourle.«.H, all the gastric
content available, likewise the Gth time, regardless of the colour. After
the 6th sample 0,oi mg of histamine pro kg body weight was injected
subcutaneously, and 10 min. after that the stomach was emptied
again. On these samples the free HCl was titrated with NaOII using
Kongo red, and the total acidity using phenolphtalcin ns indicators;
when the acidity of the sample was on the alkaline side of the change-
point of Kongo red, it was titrated with HCl until the change-point.
The results arc reported in the usual way in cem of O.i N alkali or
acid pro 100 cem gastric content. The sample quantities, thus attained,
were measured, and the dilution of colour in the samples was determined
using comparison scries. In a part of the samples the pH also was meas-
ured (electrometrically, with glass electrode).

In order to prevent the production of alkalosis, NHjCl was taken
in use. It was given to the subjects per os on the day before, O.i.'i g pro
kg body weight, divided into three portions, which dosage seemed to
reduce sufficiently the alkali reserve of the blood, so that no alkalosis
developed diuring the experiments. Tliis was made sure by determining
the pH of the subjects’ urine before the test-meal, during it and after it.
According to BRAS.SFiEr,D and Behrmann (1941) the pH of the urine
reacts liably to that pH-rise of blood which is a result of hypen’cntila-
tion caused by Oj-dcfiency, and rises about 1 pH-degree.

The oxygen deficiency experiments were performed in a lowpressurc
chamber. The subjects swallowed the stomach-tube at normal baro-

K. HARTIALA AND M. KAKVONEN.

metric pressure — moimtain-disease causes nausea, -whicli makes
the swallowing in low pressure very tmpleasant — after which the
first sample was taken immediately, and the pressure in the chamber
was brought down in a few minutes to the pressure desired in each
experiment. The low pressure was kept unchanged during the whole
experiment. — To avoid the possible disturbing effect of acclimatiza-
tion the experiments were performed during many weeks with irregular
intervals and in variable order; for the same reason it was controlled,
that the haemoglobin percentage of the subjects’ blood remained con-
stant during the experiments.

Each subject went through the following experiments:

(1) test meal in normal barometric pressure, in which the oxygen
pressure in inspired air is 149 mm, when the air is saturated in 37°C
with water vapour;

(2) as above after giving IvHiCl;

(3) and (4) as (1) and (2), but in 462 mm barometric pressure, cor-
responding to the height of 4000 m, oxygen pressure being 87 mm;

(5) and (G) as (1) and (2), but in 378 mm barometric pressure, cor-
responding to the height of 5,500 m, oxygen pressure being 69 mm.

In the 87 mm Oj-pressure all the subjects were more or less cyanosed
and did not feel very well. In the 69 mm Oj-pressure all of the subjects
were greatly cyanosed and felt seriously ill. In both of the Oo-pressures
the subjective condition was better after Is HjCl-ad ministration than
■without it. Noticeable differences between various subjects were not
observed.

Besiilts.

During the experiment it was observed, that free hydrochloric
acid was secreted in the stomach of all the subjects. The emptying
of the stomach on the subject M. K. was, as concluded from the
small amounts of secrete received and the rapid dilution of the dye,
more rapid than that of the subjects K. H. and T. L. After the
alcohol irritation the acidity values on M. K. in all the experiments
reached their maximum during the experiment, but the values
on both of the others continually rose until the interruption of the
experiment. — The variations in total acidity, HCl and pH were
parallel to such an extent, that it makes almost no difference' which
of them is followed.

The acidity values of the first sample have not any special im-
portance in our experiments, because the subjects were brought
to Oj-deficiency only after taking the sample. The other acidity
values with empty stomach before the irritant alternated relatively
irregularly, so that no conclusions can be drawn from them about
the possible influences of oxygen deficiency.

INFLUENCE or ANOXIA.

TiUtlc.

Thr Effect of Anoxia on JlCl-sccrclion.

j Without Xli,Cl.!ttimini>itrntion

-After XlbCl-iximinittr.otion ~j

llio mnxiniAl vnliio?

Till' vnhlc<
after

hwt.Mniric

injection;

r.i fllCI

Suftjfci

The nuixirnn! values

Tilt' vnliic.'?

Aflor

alcohol

initixtion;

TAfJia

’rime in

mir.tiUf

nft>-7

alcohol

(’J'.A/'HC!)

jnrt

After

alcohol

irtUation;

TA inc.l

'J'imn in
niiniitrf
after
alcohol
(TA/Hf'l)

after

lii<.'famine

injivlion;

TA jllCl

33.2, It?.:

Gn;i}0

K. //.

70.O.M9.O

tV) GO

8G.5pr..5

17.</10.i>

•10, 30

32.'i/lB.n

23,5 -■ o.e

r.ttr>o

2rt.n'i:i.5t

2].,VJ.3.i

GO/CO

2G.IV17..’.

Git

T. 1.

M'.'

]7.;i 11.5

go;go

23.7/M..S

GO,f.O

fM.IgW.S

2s.')/22.:

3'>.S/2.>.f'

3',)..V12.2

37..5/2r..o

14.r.' G.o

t'l't'-.O

37.5;2r>.o

12.5/ 7.0

cn.GO

IG,.'./ G.o

O.n/ -1.5

CO /at

l'J.fi/18.0

■to;.iO

.'iS.'.vrtO.T

A’.

lift

GG.2/i7.i;

riOoit

o-l.s/lD.i

•10/10

■1G..5/]8.0

.87

82.2 'll!.;!

•10;'3fl

.sft.i .‘lO.i'i

3S.r./l-}.n

1000

:g).i/ 7.5

G'.t

3s.5'10.j

2a.T>0

G'i.r./',0,5

The most ini})ortnnt vnlvies, obtained in the ('.xperinionts, arc
collected into the table on this page. It ])iesent.s .side by side
the grcate.st vniuc.s of total acidity and IKM without NIf,Cl-ad-
mini.st ration and after it with alcohol ns irritant, and the time that
clnp.s’c.s before the highest valuc.s arc nttnineil after giving alcohol;
when the ina.virniitn liad not been p.a.ssed during the o.xpcriinent,
it was marked GO jniiuitc.s. The third pair of columns present.s total
acidity and HCI-value.s 10 mimite.s after injecting histamine.

(1) When comparing the re.sults of acidity values in the differ-
ent 0;-prc.s.sure.s after XIf,Cl-admini.st ration and without it, it is
noticed, that there are no regular and essential differences either
after alcohol or after histnwinc irritation.

(2) When the greatest acidity values received after alcohol irri-
tation arc examined, it i.s noticed, that:

y (a) in 87 mm 0.-prcs.sure there were no e.ssential and regular
changes; in a part of the exjmrimcnts the values were higher than
normally, in a part lower, in the majority, however, lower;

(h) the acidity values in G3 mm Oj-pres-surc were regularly de-
creased in a considernble degree, compared both to normal values
and to tho.se attained in 87 mm Oi-pro.s'snre. Tree liyclrochloric
acid was to he found in all the subjects, however.

K. UARTIALA AND M. KARVONEN.

(3) Tlie acidity values wliicli were derived after histamine irri-
tant show, that

(a) in 87 mm Oj-pressure there were no regular variations; in
a part of the experiments the acidity values had risen, in another
decreased, when compared to normal values;

(b) the acidity values in 69 mm Oj-pressure had as a rule de-
creased.

Because the experiments were interrupted, it is not known, how
great the maximal acid secretion, produced by histamine, has been
in the different experimental conditions. Accurate conclusions can-
not be drawn from the fact, Avhether the oxygen deficiency in
our experiments has decreased the HCl-secretion, caused by the
injection of histamine. The subject M. K., whose reaction type was
the fastest, had not very low acidity values after the histamine
injection even in 69 mm O.-pressure. In all the experiments the
histamine injection increased the acidity values from the level they
had reached after alcohol irritation.

The curves of the subjects T. L. and M. K., presenting total
acidity both without NHjCl-administration and after it in differ-
ent Oo-pressures, give a clear idea of the influence of anoxia on
HCl-secretion. (Figures 1 and 2.)

On the subject M. K. the maximum of HCl-secretion was at-
tained earlier both in 87 mm and 69 mm Oj-pressures in accord-

Total acidji-y

INFLUENCE OF ANOXIA.

ancc with the results of Haautmann, Hei’F and Ltift. Evidently
it does not mean only shortening of the secretion time, for, wlicn
inspecting the secretion curves, it is noticed, that the HCl-secre-
tion occurred also more rapidly in normal condition.s: the ascent
of the curve is steeper. On the other subjects such shift of the
maximum is not to be found in our experiments — very possibly
due to the relative slowness of their HCl-sccrction compared to the
arrangement of our experiments. Even though the acidity values
of the subjects K. H. and T. L. continually increased, it is very
likely, that the reduction of the acidity values in their anoxia
experiments is not a consequence of the retardation, but expressly
of the decrease of the secretion. This is shown by the fact, that the
slope of the graphic curves indicating HCl-values is already before
histamine injection even on these subjects ns a rule decreasing,
mo.st apparently in the anoxia experiments. It may be mentioned,
that Hellebranjit, Broguon and Hoofes in their experiments
observed individual variations to some extent of the same type in
the reactions of the subjects’ HCksccretion to On-deficiency.

The dilution of the dye may be regarded as an indirect measure
of the quantity of the secrete in so far that, if the colour does
not weaken, no secretion takes place; if, on the contrary, it becomes
diluted, the quantities, both of the fluid secreted in the stomach,
and that flowing from it, influence the rapidity of the dilution.

K. HARTULA AND M. KARVONEN.

Depending on the abundance of the fluid, used as irritant, the
amount taken as samples, has as such no significance to the ra-
pidity of dilution, when, on the other hand, the influence of the
quantity of fluid that passes the pylorus into the intestine is very
essential. The dilution of the dye on each subject was in normal
conditions and in 87 ram Oj-pressure essentially just equally rapid.
In 69 mm Oj-pressure on the other hand the dilution on K. H. and
T. L. was very much slower and the secretion correspondingly
scantier. On the subject 51. K. the dilution was on the contrary
more rapid than in normal barometric pressure. This was certainly
partly due to the faster secretion, but another reason may also be,
that the stomach was emptied more rapidly than in normal con-
ditions. According to van Liebe Oj-deficiency — it is true —
delays the emptying of the stomach, but he mentions, that rather
great individual variations can be observed in this influence. On
the whole the dilution curves of our subjects are parallel to the
corresponding acidity value curvesj even the dilution does not
make any essential difference between the experiments without
NHiCl-administrarion and after it.

The graphic curves in fig. 3, showing the data on the subject
T. L., presents clearly the influence of anoxia upon dilution speed
of the dye.

The average of the urine-pH on the different subjects was about
6.67 in the experiments without NHiCl-administration in the an-
oxia experiments, when the corresponding average after NHiCl-

INFLUENCE OF ANOXIA.

administration was only pH 5.14. In the experiments preceded by
NHiCl-administration the pH rose during anoxia in 87 mm 0,-
pressure about 0.58 pH-dcgrees, and in 09 mm Oj-pressure it fell
0.11 pH-degrees; in the experiments without NHiCI it rose cor-
respondingly 1.16 and 0.53 pH-degrees. Consequently, NH 4 CI-
administration can be regarded as an effective antagonist to the
production of alkalosis in our experiments.

Discussion.

Earlier investigators have usually regarded the decreased HCl-
secretion during anoxia as a result of the simultaneous variations
in the ion-equilibrium of blood. It is known, that the disturbances
of the ion-equilibrium of blood for some other reasons also lead to
variations of the HCl-secrction. Bakaltschuk supposed, that the
increasing of the HCl-secrction, produced b}’’ CO.-inspiration, is
a compensative reaction, effected immediately by aciditj% wherein
the organism tries to get rid of an excess of acid even in this way;
the decreasing of the HCl-secretion, caused by vpluntary hyper-
ventilation, was to be understood correspondingly (Delhougne).
Bkowne and Vineberg regarded the CO.-content of blood as
decisive. The variations of the pH of blood or the CO.-content of
it, produced in other ways, . however, have not a similar effect
on the HCl-secretion, which was explained by Delrue and Lac-
QUET (1934) so, that the variations in question of the HCl-secre-
tion depend both on the contents of the Ca+’*' and HCOj^-ions
in blood; only the increasing of one or the other or both of them
would make the increasing of the HCl-secretion possible and vice
versa. The concentrations of the Ca"^"^ and HC 03 “-ions again
are correlated, but also connected to the concentrations of H'*'-
and HPO 4 -ions in a way, which is nearest represented by the
following equation, introduced by Kugelmass and Shohl (1924);

[ C ^;1 -x[HC^]x[HP04-] ^

[H+]

ScHiFFLERs (1936) and Thiele (1937) again are of the opinion,
that Ca+'*'-ions alone have importance to the HCl-secretion. —
The Ca+'^-content of blood is decreased in alkalosis, caused by
hyperventilation and increased in C 02 -acidosis, so that even these
theories would satisfactorily explain the influence of hyperventi-
lation and CO 2 on the HCl-secretion. On the other hand, however,

K. HARTIALA AND JI. KARVONEN.

an altogether opposite standpoint has been presented: Grant
(1941) declares, that the Ca-salts have an inhibiting influence on
the HCl-secretion, produced both by nervous and chemical irri-
tation.

It is knovn, that hyperventilation, caused by Oj-deficiency,
increases the pH of blood and decreases both the CO 2 - and Ca+’’‘-
contents of it. These variations have apparently taken place even
in our experiments. NHiCl-administration, on the contrary, de-
creases the pH of blood and increases its Ca'*”^-content, but it has
a decreasing influence on the CO.-content of blood, just as O™-
deficiency does. In our experiments NHjCl cannot be noticed to
have any essential decreasing or increasing effect on HCl-secretion;
thus the decreasing of the HCl-secretion is not effected by any of
these variations in the ion-equilibrium, but most likelj^ by the
direct influence of anoxia.*

It is not impossible, however, that the decreasing influence on
the HCl-secretion of both the alkalosis, caused by voluntary
hyperventilation, and anoxia is based — at least partly — on the
same mechanism. According namely to Campbell (1941) the 0.-
pressure in the tissues decreases, when the pH of blood increases,
e. g. as a result of hyperventilation, occurred in normal barometric
pressure. This decrease again is a consequence of the increasing
affinity of haemoglobin to oxygen, when the pH rises, and the
blood less readily transfers oxygen to the tissues. (The dissociation
curve of Oo-haemoglobin moves to the left.) Thus the influence of
alkalosis could perhaps be founded on anoxia — and not Auce
versa.

According to Mosonyi, Gunther and Petranyi the increasing
effect of CO 2 on the HCl-secretion is directed to the secreting

* After liaving finished the writing of this study, we received a publication —
delayed because of the war — in which Stajutli and Endtxer (1944) report
investigations concerning HCl-secretion in the mountains. It is interesting, that
they too have investigated the influence of previous NH,Cl-administration on the
HCl-secretion. Still, they have observed, that NH,C1 prevents the decreasing of the
HCl-secretion in the height. — The controversy between their results and ours is
perhaps only apparent, for they have made their investigations in the height of
only 3,457 m, in which height the O.-deficiency is compensated practically com-
pletely, at least at rest. The compensation happens through hyperventilation —
and leads to a transitory alkalosis. In these conditions the beneficial effect of NH,C1
on gastric secretion is a double one. In increasing tlio hyperventilation it furtliers
the compensation of the O.-deficiency. By its acidifying action, on the other hand,
it eliminates the alkalosis and with it the other possible cause of a decreased HCl-
secretion. — In their investigation the test ireals after !NH,Cl-administration and
without it are made on different subjects, which also makes the judgement of
their results difficult ns to our main question, for the individual variations in
the HCl-secretion are rather groat.

INFI.UENCE OF ANOXIA.

mucous membrane. Delkue and Lacquet also arc of tlie ojiinion
tliat tlie Ca‘’'‘*'-ions have a direct role in the reactions of HCl-
production. Our own experiments do not cxidain, Avhat the effcc'
tive mechanism of anoxia on HCl-secrction is, but it is quite prob-
able, that the influence of anoxia also acts dircetty to the secretory
cells.

The formation of hydrochloric acid from the NaCl is an
endothermic reaction, that binds energy in plenty. The source of
this energy is — so it must be supposed — one or other oxidative
reaction in the secretorj- cells. It is very probable, that the HCl-
sccretion must diminish, ns the supply of oxygen becomes difficult
and the supposed energetic reaction is reduced — whatever the
mutual amounts in blood of ions, participating in the HCl-
formation itself, may be.

Summary.

(1) The aim of this investigation is to unravel, whether anoxia
as such has — independent of the accompanying variations of
the ion-equilibrium of blood — any effect on the ga.stric HCl-
socretion.

(2) For this purpose tcst-mcals were given on throe human
subjects in a low pressure chamber, both after NH,Cl-ad ministra-
tion the day before, and without it, the Oj-pressurcs of 149, 87
and G9 mm existing in the inspired air, corrcsj)onding to the heights
of 0, 4,000 and 5,500 metres.

(3) The following results were obtained;

(a) HCl-sccretion in G9 mm O.-pressurc was greatly reduced;

(b) HCI-secretiou was affected in 87 mm Oj-pressure too, but
in a less degree and somewhat irregularly;

(4) The answer to our question seems to be — basing on the
results — that the decrease of the HCl-secrotion during severe
Oj-deficiency is a consequence of anoxia, and lia 2 )pcns indepen-
dent of alkalosis produced by hyperventilation.

K. HARTXAIA AND M. KABVONDN.

Beferences.

Bakaltschuk, M., Klin. Wschr. 1921'*. 7: II. 1551.

Bateux, R., C. R. Acad. Sci., Paris 1911. 152. 396.

Brassfield, Ch. R., and V. G. Behrmann, Amer. J. Physiol. 1911.
132. 272.

Browne, J. S. L., and M. Vineberg, J. Physiol. 1932. 75. 315.
Campbell, J. A., Physiol. Rev. 1931. 11. 1.

Delhougne, F., Klin. Wschr. 1927. 6: 1. 804.

Delrde, 6 ., Arch. int. Physiol. 1934. 38. 126.

Delrue, G., and A. Racquet, Ibidem. 1934. 39. 295.

Douglas, C. G., C. R. Greene, and F. G. Kergin, J. Physiol. 1933.
78. 404.

Grant, R., Amer. J. Physiol. 1941. 132. 460.

Hartmann, H., G. Hepp, and U. C. Luft, Luftfahrtmedizin 1941. 6. 1.
Hellebrandt, F. a., E. Brogdon, and S. L. Hoopes, Amer. J. Physiol.
1935. 112. 451.

Kugelmass, I. K., and A. T. Shohl, J. biol. Chem. 1923. 58. 649.
JIaclagan, R. F., j. Physiol. 1935. 83. 16P.

Mosonyi, j., L. GOnther and J. PetrAnyi, Z. ges. exp. Med. 1935.
95. 670.

Pickett, A. D., and E. J. van Liere, Amer. J. Physiol. 1939. 127. 637.
ScHiFFLEBS, L., Arch. int. Physiol. 1936. 43. 452.

Sleeth, C. K., and E. J. van Liere, Proc. soc. exp. Biol. N. Y. 1936.

35. 208. Quoted from van Liere (1941).

Stampfli, R., and B. Endtner, Helvet. Physiol. Pharmacol. Acta 1944.
12. Suppl. III. 189.

Thiele, W., Klin. Wschr. 1937. 16, 1. 165.

VAN Liere, E. J., Physiol. Rev. 1941. 21. 307.

Warren, C. B., Geogr. J. 1937. 90. 126.

From the School of Dentistry and the Zoophysiologicnl
Laboratory, Copenhagen.

Bficrodeterminatioii of pH in Saliya.

BODIL SCHMIDT-NIELSEN.

Received 17 October 1945.

The method hero to be described is designed for determinations
on freshly secreted saliva from the parotid and mandibular glands
respectively. The separation is carried out as described below.

Saliva contains free carbon dioxide. 'When saliva is exposed to
the air CO* is quickly lost, and the pH increases. In order to
prevent the loss some investigators (c. p. Becks, 1924, AVhite
and Bunting, 1936, Tenenbaum and Karshan, 1937) have used
paraffin oil as a protecting seal during the collection of saliva.
Linderstrom-Lang and Holter (1942) have shown, however,
that the rate of COs diffusion tlurough paraffin oil is only four
times slower than in water, and consequently the oil seal is not
very effective as a protection against escape of CO*. Other in-
vestigators have made little or no effort to prevent the COj loss.
Most of the data in the literature for the pH of saliva are there-
fore, probably, too high.

Instead of paraffin oil the author used mercury as protecting
seal. If saliva is sucked directly from the mouth into a syringe
pipette, and the rest of the space of the syringe filled up uath
mercury, no COj can escape.

Another source of error in determinations of the pH of the
normal resting saliva is caused by variations in the composition
of the secreted saliva due to the manipulations in samphng. If
the collection of the saliva sample causes any stimulation of the
glands, the pH of saliva increases markedly (as shown in the fol-
lowing paper). It is therefore important to choose a method

7 — i6021S. Acta phys, Scandinav. Vol. 11.

BODIL 6CHMIDT-NIELSEN.

that requires as little saliva for the determination as possible,
vdthout reduction in the accuracy.

The colorimetric methods for pH determination are not very
accurate especially not for microdeterminations. In applying
electrometric methods the following electrodes are available: 1.
the hydrogen electrode, 2. the quinhydrone electrode, 3. the glass
electrode and 4. the antimon electrode. It is obvious that the
hydrogen electrode with bubbling hydrogen cannot be used, be-
cause the bubbling through the saliva wiU wash out the COj.
Other forms of hydrogen electrodes could perhaps be used,
but the use of hydrogen is always inconvenient. The glass
electrode requires too much saliva. The ordinary forms require
about 5 ml fluid, and it will be difficult to prepare an electrode
that wiU require less than about 0.5 ml. The antimon electrode
was used by Stephan (1940) for determinations of pH on the
surface of the teeth with very satisfactory results. It seems that
this electrode, in connection with a calomel bridge direct to the
mouth, is very well suited for pH determinations inside the mouth.
However the antimon electrode is not very accurate, so that when
the determination can be carried out outside the mouth the quin-
hydrone electrode will be more favourable.

Biilmann’s capillary quinhydrone electrode is in many re-
spects well suited for this special purpose, a) The volume of saliva
necessary for a single determination is only 0.08 to 0.02 ml.
b) The method is accurate when the pH to be determined does not
exceed 8. The pH of saliva never exceeds this value, c) The pres-
ence of proteins does not produce any measurable error when the
protein content of the sample is below 1 — '2 per cent. In saliva
the protein content is about 0.5 per cent, d) "When the concentra-
tion of salts is below 0.2 — 0.5 n, the salt error is negligible. The
concentration of salts in saliva is about 0.02 — 0.04 n.

Method.

Apparatus-

1. Tivo rubber cups as used in Gore’s saliva separator. Big. 1.
When the rubber cups are applied to the mucous membrane sur-
rounding the papillae of the parotid ducts the parotid saliva will
drop out of the glass tubes like saliva from a Pavlov fistula in
a dog.

pH vs

_ ate coveting

less steel [^)J^J^ opposite tO

X. «-

tlie syiing®’ ® to ^^to tbe syxiiig®- ^

*'tf-l»:ieTec«ae.

5. A potentiometer.

Ptooedore. ^otii o!

torn aps®e.“*” “4 eomeotei w& S

t«mm cUoni® and com

100

BODIL SCHMIDT-NIELSEN.

Scm

p— </

The syringe is waslied -vrith distilled water and the dead space
filled with a suspension of quinhydrone in a httle water. About
0.05 ml of a standard buffer solution (pH 6.00 or 6.50) is sucked
into the syringe and the rest of the space filled up by mercury.
The mercury fills out the dead space and a drop of mercury hes
in the lumen of the syringe. When the syringe
is shaken thoroughly the mercury drop will serve
for a complete stirring of the content of the
syringe. AVhen the stirring is finished, the syringe
is held in a vertical position and the mercury is
driven out, and now the mixture of buffer solu-
tion and quinhydrone shall be transferred to the
electrode. The plane end of the small glass tube
from the electrode (a. fig. 2) is pressed against
the tip of the syringe and the mixture is driven
out into the tube. The electrode is put together,
placed in the holder and the adjusting of the
potentiometer takes place.

WTien the potentiometer is adjusted to the
standard solution the determination on sahva can
be carried out. The sjrringe is washed and the
dead space filled with the suspension of quin-
hydrone. The saliva samples are taken in the
following ways.

The parotid saliva. The person is sitting with
the rubber cups (mentioned above) covering the
papillae of the parotid ducts. The parotid saliva
is dropping out of the glass tubes. The sahva
nearest the outlet of the tube is sucked away
with a piece of filter paper, the tip of the syringe
is then pressed against the tube and about 0.05
ml sahva sucked into the syringe.

The mandibular saliva is sucked directly into
the syringe from the floor of the mouth just
under the tongue. In order to prevent loss of
CO- the mouth must be kept shut until the
moment when the sample is taken.

In a fohowing paper the results of a series of determinations
of the pH of normal resting sahva from the parotid and mandib-
ular glands wih be pubhshed. (All the measurements in this
and the fohowing papar are carried out at room temperature.)

Fig. 2. A Bul-
MAM? capillary
electrode,
a. Glass tube. b.
Platinum wire.
0 . Rubber stop-
per. d. Contact
for connection
with potentio-
meter.

MICRODETERMINATION OF pH IN SALIVA.

101

Experimental.

Some , experiments to demonstrate tlie accuracy of the method
are given here.

Table 1.

Determinations on saliva samples with a quin~
hy drone electrode and a glass electrode.

Sample no

Quinhydrone

electrode

Glass

electrode

7.00

7.01

7.07

7.17

7.15

7.22

7.28

7.29

7.42

7.48

In table 1 the results of pH-determination on different samples
of saliva are given. The determinations are carried out both -with
a quinhydrone electrode and a glass electrode. The agreement be-
tween the results obtained with the two electrodes is seen to be
satisfactory. (The results are higher than usual for saliva samples,
but this is due to the fact that all the samples have been exposed
to the air, and furthermore some of the samples have been col-
lected during stimulation of the secretion.)

The effectiveness of a protecting paraffin oil seal has been
tested in the following experiment: A saliva sample is divided into
two equal parts each about 0.5 ml
and placed in two small cylindrical
vessels of 2. 5 cm diameter. One of the
samples is covered with a paraffin oil
seal 5 mm thick. pH is determined in
the two samples every ten minutes.

Erom the results in fig. 3 it is seen
that the oil seal affords a certain
protection. During 50 min. pH only
increases 0.18 pH units in the pro-
tected sample while pH in the other
sample increases 0.32 pH units. But
the fig. does not show the increase in
pH during the collection of the saliva
sample. The usual mode of proceeding

pH
760
7.5o
740
7Jo
7^0
Zlo
7.00
CSo

10 10 JO 40 SO so min.

Fig. 3. pH variations of sample
protected with pMaffin-dil-(-dO-c::^t-
and unp^otected'Bample*^(••/. (J

102

BODIL SCHMIDT-NIELSEN.

for pH determinations of former investigators has been to let
the saliva drop down from the mouth into a vessel containing a
little paraffin oil. During this process some CO 2 is lost. In one
experiment a sample was collected in the manner described and
pH determined. The value found was 7.01. A determination on
the saliva taken by the help of a syringe directly from the mouth
of the same person gave 6.72.

Aooj — — — _ _ —

7.90 - ^ ...

7tfo - •

7.70 - •

760 -

7.5o - •

?4o -
7.3o .

7^0 -

7.I0 - •

7oo

•

d9o .

X « « - » ‘ I i I - t_ t I i_ _i_ 1 _t i -

12 3 4 5 C 7 B 9 10 11 12 13 1* 15 ft min.

Fig. 4. pH variations of saliva sample stirred in an open vessel.

An experiment has been carried out in order to see how much
the total loss of COj influences the pH of a saliva sample. A saliva
sample is placed in a small open vessel and pH determined. The
sample is stirred for 10 sec. and pH determined again. The pro-
cedure is repeated until constant values are obtained. The results
are given in fig. 4. It is seen that after a total of 8 min. stirring
pH reached a constant level. The corresponding increase in pH
is one pH unit. But 50 % of the increase takes place in the first
one and a half minute of stirring.

A series of consecutive determinations on two standard buffer
solutions and a saliva sample have been made in order to test the
accuracy of the method. (To avoid increase in the pH of the saliva
sample during the determinations, the sample was stirred a Httle
before the determinations and covered by paraffin oil.)

JIICRODETERMINATION OP pH IN SALIVA. 103

The results arc given in table 2. It is seen that the accuracy
is about 0.01 pH units.

TnWo 2.

Consecutive determinations on two standard
buffer solutions and one saliva sample.

Standard bufibr
pH 6.50

standard buffer
pH 7.50

Saliva sample

6.50

7.40

7.08

6.50

7.49

7.08

6.50

7.48

7.09

6.50

7.49

7.08

6.50

7.48

7.09

6.50

7.48

7.10

Summary.

A method is reported for determinations of pH in saliva from
the parotid and mandibular glands. A micro quinhydrono elec-
trode according to Biilwann is used. A technique is described
by vhich loss of COj during both collecting and measuring of
the sample is avoided.

Determinations ate carried out to show the accuracy of this
method in comparison with methods used by former investigators.

I wish to express my most hearty thanlcs to dr. Holder
Jorgensen, dr. H. Ussing and dr. B. Asmussen for valuable
help and suggestions during the development of the method.

Referoncos.

Becks, H., Ergebnissc neuerer Speichel-Untcrsuchungen. Berlin 1929.
Bhlmann, E., Bull. Soc. chim. France 1927. 41. 213.

Gore, J. T., J. dent. Res. 1938. 17. 69.

Keogh, A., Ind. Eng. Chem. 1935. 7. 130.

Linderstrom-Lang, K. and H. Holter, C. R. Lab. Carlsberg, Ser.
chim. 1942. 24. 105.

Stephan, R. M., J. Amer. dent. Ass. 1940. 27. 718.

TENENBAXJSt, B., aud M. Karshan, Ibid. 1937. 24. 1255.

WHITE, J. and R. W. Bunting, Amor. J. Physiol. 1936. 117. 529.

From the School

The pH in Parotid and Mandibular SaUra.

BODIL SCHMrDT-lOELSEN.

Received 17 October 1945.

the pH iCoM deteimieafion of

JTielsek, 1946) A se^™ f ™s described (Sohmibt-

i. «. ->2: -

obtLe°d tahva caa be

and'd^C\“!r “ during a da,

pe^om. “ ""*■”8 aaHva from a number of

in the hterature whorf^ <ieter^ations in sahva are available
terogeneous. This is ^

Biethods for coU ctiof of^or inadequate

loss from tL T"^i n determination of pH, CO,

single one of these^acf of the secretion. Every

the results. ^ produce large variations in

the authorslift^^riven too qnaUty of the results, because

and the experimental condi w' ormation about the procedure

-ain p^IroVt"-^ determinations were incidental to the
considered their choice^^T^ the authors have not

cases been a^od hT in a few

satisfactory.^ ’ * have been un-

105

pn IN PAROTID AND MANDIBULAR SALIVA.

Extensive lists of the older literature are given by Babkin
(1928) and Boseman (1927 and 1932). A list of the newer in-
vestigations is given by Bazant and Studni6ka (1941),

min.

Fig. 1. An example of the increased
pH in the beginning of an experiment.

• Saliva from the left parotid.

X Saliva from tbo right parotid.

The Conditions nndor which Resting Saliva can be Obtained.

I have found that constant values for pH of saliva can bo ob-
tained only when the person studied is sitting completely quiet,
every sort of talking, mastication, sneezing or coughing causes
an increase in the pH of saliva. The increase depends on the in-
tensity of the stimulus and is often
remarkable. As an example can be
mentioned that the pH in parotid
saliva increased from 5.62 to 6.68
after a single short fit of coughing.

In the beginning of an experiment
the pH of saUva usually is some-
what increased (fig. 1). This is
mo^ probably due to the irritation
caused by the rubber cups covering
the outlet of the parotid ducts. The
pH soon decreases to a constant
level, and as shown in fig. 4 and 5
this level is very constant for the same person. Wlie^i measuring
the pH of resting saliva it is therefore necessary to repeat the de-
iertninalions until a constant level is reached.

As every sort of stimulation of the saliva secretion causes an
increase in the pH of saliva (shown in table I) it seems most prob-
able that the lowest attainable value for the pH in saliva from
a certain person is that nearest to the correct value for resting
sahva for that person. Also the constancy of this lowest level in-
dicates that the real value for resting saliva is attained.

Eig. 2 gives an example of how small the pH variations in
parotid saliva can be in a period of 24 minutes, when the person
is sitting very quiet. The pH is measured at intervals of 3 minutes.
At the point marked by an arrow the person was told to do
mastication movements, as seen in the fig. this causes a marked
increase in pH.

As mentioned in the preceding paper sampling of the mandib-
ular saliva involves suction by the syringe direct from the floor
of the mouth. The touching of the mucosa, however, causes a

106

BODIL SCHMIDT-NIELSEN.

stimulation of the secretion -with increase in the pH of the sahva.
The increase has appeared not to last long. But if the samples
are taken at short intervals (about 3 minutes) too high values
will bo obtained. Intervals of 10 minutes seem to be sufficient

J 6 9 tt ts IS t1 27 K srnin.

Fig. 2. pH variations in saliva from one of the parotids during a period of rest.
At tho point marked by. an arrow the person begins to do mastication movements.

to get satisfactory results. Fig. 3 gives an example of determina-
tions in mandibular saliva. The first 7 samples were taken mth
intervals of 3 minutes, the later with intervals of 10 minutes.
The results are definitely lower ■when the intervals of sampling
are 10 minutes instead of 3 minutes. The average pH in the first
samples is 6.84, in the last samples the average pH is 6.56. It is
thus necessary, xvhen pH in mandibular saliva is determined, to use
intervals between sainplings of at least 10 minutes.

S.SC

6.00

Jm/n, /Mfryafj TOmift.tnffrwU

Fig. 3. pH in mandibular saliva determined %Wth intervals of 3 minutes and 10

minutes.

The Variations in pll in Parotid and Mandibular Saliva
during a Day and during a Longer Period.

In fig. 4 the results of a series of determinations showing the
variations during a day are recorded. The determinations were
made on both mandibular and parotid saliva. Determinations
were made on saliva from both parotids, but as the two parotids

pit IN PAROTID AND MANDIIJOI.AR SAMVA.

107

Fig. 4. Tlio vnriationa during n day in mandibular (>:) and parotid (O) Ealiva

from 7 persons.

The vertical lines represent the hours of the meals.

Abscis-sa, liours of tlio day.

always give nearly identical results, the values for pH in parotid
saliva arc only marked by one point in the figure. Pivc pensons
were investigated 4 . times during a day, and two persons were
investigated 7 times during 2d hours. The hours for the deter-
minations and for the meals arc marked in the figure. The two
persons who were investigated in 2d hours. went to bed after the
sampling at 11 o’clock in the night. They rose for an hour for the
sampling at 3 o’clock in the night, after that they slept again
until just before the samiiling at 7 o’clock. From the figure it is
seen how remarkably constant the pH in the parotid saliva is
for all the persons studied. The pH is independent of the meals
and also of the time of day or night. The standard deviation for
one person during a day is about 0.06 pH units.

108

BODIL SCHMIDT-NIELSEN.

Fig. 5. Tho variations during nine month
in pH of mandibular (x) and parotid (©)
saliva from one person.

Abscissa, dates.

The pH in the mandibular
saliva does not show quite
the same constancy as the
pH in parotid saliva. (This
is probably due to the more
difficult sampling where,
as mentioned above, the
sampling itself causes an
increase in pH.)

In fig. 5 the results of
determinations on saliva
from one person during nine

months are represented. Also in this case the pH of parotid

saliva is remarkably constant, while the pH of mandibular

saliva is more variable showing a standard deviation of 0.17

pH units.

The pH in Besting Saliva from a Number of Persons.

The pH has been determined in resting saliva from 40 persons.
The results are given in fig. 6. The 19 first persons are clinical
assistants and students from the School of Dentistry, the re-
maining 21 persons are all pregnant women from the pregnancy
consultation of the "Rigshospital”.

It is seen that pH in the
parotid saliva ranges from
5.45 to 6.06 (average 5.81).
pH in the mandibular sa-
liva ranges from 6.02 to
7. 14 (average 6. 39). There is,
however, a systematic dif-
ference between the man-
dibular saliva from the
pregnant women and the
corresponding saliva from
the other persons studied.
The average for pH in
mandibular saliva from the
pregnant women is 6.22,
while the corresponding
average for the other per-

Axilitonfs and

firtgnant

Fig. 6. The pH in mandibular (x) and
parotid (O) saliva from 40 persons.

Abscissa, the persons studied.

pH IN PAROTID AND MANDIBULAR SALIVA. 109

sons is 6.67. Furthermore the individual variations are definitely
smaller for the pregnant women.

It is not possible to trace the cause of this difference from
the material presented.

The Influence of Stimulation upon the pH of Saliva.

The effect of three kinds of stimuli has been studied. These
were: gustatory stimuli, olfactory stimuli and mechanical stimuli.
Only the influence upon the parotid saliva has been studied, be-
cause the gustatory stimuli contaminate the mandibular sahva.

The procedure was as follows: First the pH in resting saliva
was determined, then the stimulation was commenced and con-
tinued, coincident with determinations of pH, until a maximum
in the increase of pH was reached. The gustatory stimuH: sugar,
citric acid, salt and quinine were apphed as crystals, which were
placed on the tongue and distributed by small movements of the
tongue.

In table 1 the results are given. It is seen that the gustatory
stimuli have by far the strongest effect, but also the mechanical
stimuli have marked effect while the olfactory stimuli have little
or no effect.

Table L

Gustatory

stimuli

rest.

sal.

stim.

sal.

Olfactory

stimuli

rest.

sal.

stim.

sal.

Mechanical

stimuli

rest.

sal.

stim.

sal.

Sweet (sugar). .

6.83

7.59

Ammonia

Coughing . .

5.02

damp.

5.81

5.88

Aeid(eitrie acid)

6.03

7.80

Formalde-

Chewing . .

n 7 0
0,1

6.34

hyddamp.

6.81

G.07

Salt (salt)

6.83

7.52

Smell of

Touching

chocolate .

ran

of the oral

mucousn .

6.82

6.90

Alkaline (so-

bonatc)

6.00

Summary.

The author’s micro method for the determination of pH in saliva
(Schmidt-Nielsen, 1946) is applied. It is found that:

1) Besting saliva can only be obtained when the person studied
is sitting completely quiet. Even talking, mastication, coughing

110

BOWL SCHMIDT-NIEI^EN.

or anytliing like that causes a marked increase in the pH of
saliva.

2) The pH of resting parotid saliva is remarkably constant for
the single person, not only during 24 hours, but also from day to
day and from month to month. The pH of mandibular saliva does
not show quite the same constancy, this is perhaps partly due
to the technique employed.

3) The pH’s of parotid saliva from 40 persons range from 5.45
to 6.06 (average 5.81). The pH’s of mandibular saliva range from
6.02 to 7.14 {average 6.39).

4) Stimulation causes a considerable rise (up to two pH units)
in the pH of the saliva. Gustatory, olfactory and mechanical
stimuli were tried. Gustatory stimuli were found to have the
strongest effect upon the pH.

Roforoncos.

Bahkin, B. P., Die iiusscrc Sekretion dcr Verdauungsdrusen. Berlin
1923. 2’Ed.

Bazant, V., and F. K. StddmCka, Das Problem dcr Vitalitiit dcs Zahn-
schmolzcs. Prag 1911.

Eose.mank, 11., Hnndb. norm. path. Physiol. 1927. 3. 819.

Rosemaxn, R., Ibidem 1932. IS. 53.

ScniiiDT-NiELSEN, B., Acta physiol. Scand.

From the Physiology Institute at Helsinki University.

Perception of Weight and the Phenomenal
Regression to the “Real” Weight
(Thing Constancy Phenomenon).

Experiments on Arm-Amputated Subjects.

EEVA JALAVISTO.

Received 23 October 19-15.

In the visual field different constancy phenomena c. g. the con-
stancy of visual size, shape, brightness and colour have been
thoroughly investigated (for .references consult c. g. Brunswik
(1934) or Thottless (1931, 1932)). Similar constancy or invariance
phenomena also exist in other sensory modabties, but in the field
of weight perception the results of the investigations are, boAvever,
still somewhat unsatisfactory. This depends partly upon the fact
that some results of the sensory muscular physiology have not
been sufficiently taken into account, partly because the dual
(tactual and proprioceptive) basis of weight perception has not
been adequately considered. It seems thus necessary to discuss the
peripheral physiology of the sensations of weight before dealing
with the particular problem of the phenomenal regression to the
"real”'- weight.

The basis of weight perception varies both in similar experi-
mental conditions between different subjects and in different
conditions regarding the same person. As a general rule indications
of the perception of weight are derived from skin receptors in all
persons without exception when the lifted weight is small and

' For the meaning of “real” in tin's conneution sco a note by Tuouless (1931).

PERCEPTION OF WEIGHT AND PHENOMENAL REGRESSION, 113

the skin would increase with the increase in the standard stimulus
weight (according to Weber’s rule) it would, perhaps, at a certain
weight level, he greater than the weight that would correspond
to the muscular, proprioceptive difference threshold. In this case
it seems quite natiural that the receptive basis shifts from tactual to
proprioceptive.

(The increase in the difference threshold would then stop until
a further rise in the standard weight brings new muscles into
action, thus tending again to raise the (proprioceptive) difference
threshold (Jalavisto (1935).)

llTiich are the phenomena of constancy or invariance in weight
perception and how do they manifest themselves? Generally speak-
ing the thing constancy can be defined in the following way: When
phenomenal equahty exists in spite of differences in stimulus con-
ditions and this equahty corresponds to a conceptual thing charac-
ter e.g. the measured weight, there is conceptual invariance of the
perception; in other words, the perception shows phenomenal
regression to the "real” object. The main interest of this investiga-
tion is concentrated on the question of the existence of such a
thing constancy phenomenon in the proprioceptive field, i.e. on the
perceptions of weight based on impulses arising in the muscles and
tendons. IVhen the 'perception of a ‘weight sufficiently great to ensure
a proprioceptive receptory basis of the sensation is not e.g. influenced
by the arm weight there would he a constancy phenomenon of this kind.
It might he necessary to mention that if the weight to be lifted
IS small and the sensation of weight thus based on impulses arising
in the sldn and this perception remains uninfluenced by the arm
weight, it certainly does not mean the existence of conceptual
thing” constancy. Earlier investigations dealing with the effect
of arm-weight on the weight perception have not made this dis-
crimination between tactual and proprioceptive cues in weight
perception. Katz (1920) e.g. stated, that a forearm amputated
person who compared weights with his intact forearm and his
amputated forearm (the moment of weight in relation to the elbow
joint being the same) found that weights of 1,000 and 1,006 g were
phenomenally of equal weight. Undoubtedly the comparison of
weights is uninfluenced by the arm-weight in this subject, but the
weight is not sufficiently great to ensure a proprioceptive nature
0 the sensations. Eischel (1926) again, when treating the same
question, made his experimental subjects compare two weights,
one of which was lifted with the arm in a water bath (the water

8— 460215. Acta phys. Scandinav. Vol. 11.

114

EEVA JAIiAVISTO.

supporting the weight of the the arm), the other weight lifted by
the other hand in an ordinary way in air. His-experiments showed
no influence of the arm-weight, but the same objection against his
method as against that of Katz is valid. The weight lifted in this
case was also about 1,000 g and the perceptual basis of a weight
of this magnitude is probably tactual as e.g. the investigations of
Wangel, Eljigken, V. Bagh and Renqvist seem to show. Other
experiments by Eischel might, however, prove the existence of
a central invariance phenomenon, but the perceptions are probably
not proprioceptive. Fischel stated that a 1,000 g weight could be
lifted in quite an arbitrary way: by teeth and arm, by one finger
and four fingers, the arm loaded or without an extra load, and so
on, but yet the weights were judged equal. Fischel explains the
fact referring to the view put forward by jILirtin' and Muller,
that the perception is based upon the impression of the absolute
weight. The meaning of this is, however, quite identical with the
statement that the perceptions show phenomenal regression to the
“real” Avcight. Panzel (1925) has made similar investigations, but
his results were partly different. In his experiments the weight
to be lifted was heaAuer, 2,000 g or more, and the muscular condi-
tions, muscular force e.g. mflucnccd the sensation of weight, but
nevertheless the arm-xveight had no effect. These experiments,
made on two subjects, were arranged as follows: The weight was
lifted alternatively with the arm flexed at the elbow joint, about
90 degrees, the weight of the arm acting against gra%dty, and
alternatively with the arm supported in a horizontal plane, the
arm-weight thus being compensated. In these experimental con-
ditions it is, however, possible that the differenee of the arm posi-
tion in relation to the shoulder joint might have influenced the
result of the experiment. The length of the muscles in these two
positions cannot be considered equal, the biceps muscle being shor-
ter in an adduced position of the arm. Thus even if the arm- weight
would tend to make the weight phenomenally heavier, the differ-
ence in muscle length (that leads to an underestimation of the
weight in the position which corresponds to the reduced muscular
length) might have compensated and masked this effect so to say.

An other case of phenomenal regression of the perception of
weight towards a conceptual, thing character is suggested by Katz
and Stephenson (1937). MTien a weight is lifted by a string, the
weight perception corresponds to a great degree to the conception
of elasticity. The sensation of weight is viz. influenced in a greater

peeoepxiok op weight and phenomenal regression. 115

degree by tbe elastic properties of tbe string than by tbe muscle
tension developed. In this case there need not be a question of a
phenomenon of objectivation (phenomenal regression to a thing
character), however, as it can be explained by the fact that the
proprioceptive sensitivity to weight differences diminishes by
shortening of the muscle. If the weight which is lifted wthout a
string is lifted from the same end position that the arm reaches
by stretching the string, no overestimation of the weight occurs.
By this arrangement two weights lifted by and without a string
are phenomenally equal (Jalavisto, Kalin, Parvela 1938).

Earlier investigations thus give no clearcut evidence of the
existence of central constancy phenomena in the field of proprio-
ceptive weight perception. It is true, that the well-known weight-
size illusion can, according to Thouless (1932) be treated in much,
the same way as c.g. visual constancy phenomena, but the recep-
tive mechanism in the usual experimental arrangements is surely
tactual. This phenomenon does thus not concern our problem.
Especially the influence of arm weight on proprioceptive weight
sensation requires further investigation. This paper deals with the
special question whether it is possible to eliminate, so to say, the
weight of the own arm, when the w'eight sensation is based upon
the intensity of the muscle tension in the Ufting muscles (per-
ceptual indications being proprioceptive) and thus really perceive
the "physical” weight. As the peripheral receptive mechanism
shpws great individual variations the experiments must be ex-
tended to a sufficiently great number of subjects. The investiga-
tion to be presented forms a part of an extensive study which
concerns the psycho-physiology of the "phantom-illusion” of the
amputated.

Methods.

If an arm-amputated person is asked whether there is any phenomenal
Iference of weight between his amputated arm and the intact arm,
e answer is sometimes that the amputated arm is lighter than the-
ealthy arm. If the amputated arm is judged lighter, it seems evident
at the weight of the intact arm influences the perception of the arm-
weight The perception of the weight of the own arm, in this case,
cer ainly shows no “phenomenal regression to a thing character” or
mvariance; the sensation is consistent with the peripheral stimulus.

• f * f*’ when both the arms, amjiutated and

m act arm, are judged equal. In this case it must be supposed, that the
m-weight is a part of the body image and as such a fixed conceptual

116

EEVA JALAVISTO

invariance and a difference or variation in the peripheral stimulus con-
ditions (diminuation of the arm-weight by the amputation) does not
affect the weight sensation. As the indications for the perception of .the
own arm-weight cannot be but proprioceptive, this dual behavior of the
amputated persons shows, that in some cases there is a central trans-
formatory constancy, but in others the arm-weight sensation is depen-
dent on differences in stimulus conditions. The result obtained in this
simple way is, however, not quite rebable. The amputated arm is some-
times judged even heavier than the intact arm. This may be rmderstood
only by supposing, that a peripheral irritation in the arm stump in-
fluences the perception of weight quite as it molds the body image so
that the "phantom arm” sometimes feels swollen, sometimes shortened
or small, hot or chilly, and so on. As it is difficult to know in which
case the peripheral irritation has influenced the perception of the weight
of the own arm, one cannot rely solely on the phenomenology of the
arm-weight of an amputated person, especially because the perception
of weight of the own arm might possibly not have the same receptory
basis as the perception of an external weight to be lifted. The question
of conceptual constancy in weight perception must therefore be treated
experimentally. The experiments are preferably to bo made on amputa-
ted persons, so that the value of the beforementioned sjTnptom: “Ughtcr”
— "of equal weight” may be settled at the same time. The result of
such an investigation may be that always when an amputated arm is
judged lighter, it may also be demonstrated that the weight of the
subject’s own arm influences the perception of an external weight and
vice versa. In this case it is probable, that a peripheral irritation in the
stump will not interfere wth the sensation of weight of the own arm.

There are two ways, in principle, of eliminating the weight of the own
arm.

1 . The weight comparison is made between a weight lifted with the
arm in a water bath, when the water supports the weight of the
arm and a weight lifted in an ordinary way, the arm in air. In the
latter case the weight must be judged heavier if there is no trans-
formatory constancy and of equal weight if the weight sensation
shows phenomenal regression to the "real” weight.

2 . The weight comparison is made between a weight lifted in an arm
position where the arm weight acts against gravity and a weight
lifted by an arm hanging right down, the tissues passively supporting
the arm.

The former experimental method is inconvenient, because, if the
•experiments are to be made with arm amputated subjects and the
weight of the arm stump is to be eliminated, the water bath must
reach over the shoulder of the subject. This is not possible with our
laboratory conditions. The trouble with the latter method is, that to
perform both lifting movements in exactly the same arm position which
is necessary to avoid an influence of the muscle length on the sensation
of weight, the subject must be tilted between lifting the two weights,
the standard and the variable weight. This again is unsuitable, because

PERCEPTION OF WEIGHT AND PHENOMENAL REGRESSION. 117

the up and down movement might have an irritating effect on the laby-
rinth. This method was nevertheless chosen, because it had some ad-
vantages compared with the former method and the irritation of the
labyrinth seemed to be quite unimportant when the tilting was done
slowly. Only 2 out of 5i persons became nauseous, the rest did not
complain of any discomfort. On the other hand the underestimation of
weight caused by an irritation of the labyrinth or tonic reflexes seems
to be rather unimportant (Allers (1909), Mann (1912), Holmes (1922),
Aibiu and Jalavisto (1939)).

A. Adjustable supporting rope.

Lever with arm cuff.

L: Pneumatic arm cuff.

D: With ball bearings fixed axis of lever B.

L. Counterbalance of the lever B.

C- 1 -no the action of weight G.

H: Soft cusHo^:^

T. tilting mechanism.

btand to stop the tUting movement.

Supportmg board for the arm.

supporting framework of the tilting mechanism is omitted.

nieohaniM^(Kg^lf Jfb standing with his side against a tilting
an angle of 70 x * x? leamng against a support which formed

20 defees in arm was flexed about

''O degrees Above abduced in the shoulder joint about

S • bove the shoulder there was in the ventrodorsal direction

118

EEVA JALAVISTO.

a bar fixed mth ball bearings and provided with a lever parallel to the
arm. In the lever there was an adjustable arm cuff and a hook for the
suspension of the weight. The suspending rope went over a wheel, so
that the weight always pulled the lever in a right angle. In an opposite
direction there was a supporting rope which prevented the weight from
pressing the arm between the experiments, when the arm was held
passive. The length of the rope could be adjusted so that immediately
when the patient abduced his arm, the weight came to exert its whole
weight on the arm. The same apparatus was suited for experiments
both on right and left arm: the patient had only to turn 180 degrees
round his vertical axis, and so the apparatus could be used for experi-
ments with the other arm too. The inside of the arm cuff was bolstered
up with a pneumatic rubber bag to ensure a proper fitting for every arm
size. The soft pneumatic arm cuff is necessary also to avoid sensor)-
impulses from the skin and deeper tissues to influence the weight sen-
sation, thus favouring the proprioceptive basis of sensation.

The experiments were performed in the following way:

a) The main experiments: The standard stimulus weight (3,000 g) lifted
vdth the patient standing upright, the arm leaning towards a sup-
port, the arm 70 degrees abduced. AlTien lifting the variable weight
to be compared the subject lay in a tilted position of 70 degrees with
his arm hanging right do\vn. The subject has to judge whether the
second weight (= standard weight *idditional weight varied in
steps of 100 g) lifted in tilted position is heavier, equal to or lighter
than the standard weight lifted in upright position. An additional
weight which in % of the cases gave the judgement “heavier” was
considered as the difference threshold weight in these conditions.
The experiments were performed both on the amputated and the
intact arm.

b) As control experiments, determinations of the difference threshold
weight with the 3,000 g standard weight were made with the stan-
dard weight and the variable weight both lifted in the same erect
posture of the subject, the arm abduced as in the main experiments.
These control experiments too were performed with the intact as
well as with the amputated arm. As a rule every additional weight
(100, 200, 300 etc.) was tried ten times in the a) experiments, six
times in the control difference threshold b) experiments.

The experiments were carried out on 27 arm and 27 forearm ampu-
tated war invalids' operated in the years 1941 — 42 and examined
from, 1 — 12, most 8 — 9 months after the amputation. The number
of the right- and left-hand amputations was in both groups nearly
the same: among the arm amputated there were 14, among the forearm
amputated 15 right-hand amputations. For technical reasons only per-
sons with nearly the same and not too short a length of the arm stump
could be selected for the experiments. Among the forearm amputated
the differences in stump length were somewhat greater. The stump

' Patients at the Invalid Hospital of the Finnish Red Cross.

PERCEPTION OP WEIGHT AND PHENOJIENAL KEGRESSION. 119

length was measured in all subjects, but the differences were not as
great as to justify a grouping of the cases and separate treatment ac-
cording to the stump length.

All the subjects were asked, in addition to the weight discrimination
tests, whether the phantom-arm or the stump was phenomenally heavier,
equal to or lighter than the intact arm.

Results.

Control Experiments. Weight discrimination in upright posture,
the arm weight acting against gravity. In the first place we had to
test the difference threshold with a standard weight of 3,000 g
in these experimental conditions. In some persons there was
a so called negative time error, i.e. the same or even a hghter
weight lifted after the standard weight was judged heavier. In
these cases it happened occasionally, that the second variable
weight had to he reduced to obtain perceptual equality. The dif-
ference threshold weight is then marked = 0 or negative. No
difference between the stump or the intact arm concerning the
magnitude of the time error could be observed. Fig. 2 shows the
distribution of both the positive and negative threshold weights
in different experimental groups. The ordinate represents the
number of cases, the abscissa the difference threshold in grams.
In the histogram the difference threshold weights in the experi-
ments with the intact arm, the arm- and the forearm-stump are
separately indicated. As can be seen the distribution of the differ-
ence threshold weight in aU these three groups is quite similar.
In every group the difference threshold weight is usually about
-f 100 g. This seems to indicate, that on the basis of difference
threshold experiments the arm weight seems not to influence the
weight perception. In Fig. 3 the differences between the threshold
of the intact arm and the stump arm experiments are separately
plotted. A negative value indicates, that a difference threshold
weight obtained in stump arm experiments is greater than that
obtained in the intact arm experiments on the same subject.
If the weight of the arm were to influence the magnitude of the
difference threshold, the histogram would be unsymmetrical, so
that the positive values would be more numerous than the negative
ones. The arm amputated cases would also tend more towards
positivity than the forearm amputated cases. A tendency in this
direction seems to exisv between the arm and the forearm ampu-
tated cases, hut the difference is too small to be of any statistical

®'& nificance . '^■^^avibto

®«^cu;ate(J "^■7, ivjjen .

I “Oovo = t®-

'■’'■» of «,„.?“« hfer „„ ;. °'''

‘"‘ ‘io guest,™ eVuL®'™”* Wit ,> ““ "-“"M

ruie ief^ tiafcin? ®<^Wedfn I, de-

j^®^QVisr, V j{ (MfJ This

PERCEPTIOir OF WEIGHT AND PHENOMENAL REGRESSION. 121

subjects is so often greater than in tbe intact arm experiments,
shows, that there is no reason to believe that a possible influence
could be masked by the inaccuracy of the threshold determination.

"Weight discrimination in uf right and tilted fosture, the arm weight
alternatively compensated. Before conclusions can be made from
experiments in which the subject will repeatedly be tilted, the
question must be settled to what degree and particularly in what
direction the possible irritation of the labyrinth may influence

Kg. 3. Differenco of the threshold weights in intact arm and stump arm experi-
ments. Ordinate: number of cases, abscissa: difference threshold weight of the
stump experiments substraoted from the intact arm difference threshold weight.

(Black blocks: forearm stump, white blocks: arm stump experiments.)

the results. As the labyrinthine irritation in two persons was so
strong that they became somewhat nauseous, it may be supposed,
that the weight sensation in these subjects would be more altered
than in those ones who did not complain of any discomfort. Both
the nauseous subjects were overarm amputated cases. The results
of the experiments performed on the arm stump are in these two
cases opposites. In one of these subjects the weight to be lifted
with the arm stump in the tilted position was lighter than the one
lifted in the upright position. In the other, however, the weight
lifted in the tilted position was phenomenally heavier, i.e. contrary
to the expectation if the weight of the stump would affect the

122

EEVA JALAVISTO.

■weight sensation. The difference threshold was consequently
negative and one of the biggest negative values ever encountered
even in the stump experiments. The change in position and the
possibly resulting labyrinthine excitation could consequently at
most diminish the effect of the arm-weight, but certainly not
enforce it. A similar negative difference threshold was, however,
sometimes met with in control experiments in which there is no
question of a Iab)T:inthine effect, thus showing that the negative
threshold in the nauseous subject may not have any significance.

Kg. 4. Distribution of the “weight difference” in intact arm experiments (blaek
blocks) in forearm stump e.xperiments (white blocks) and arm stump experiments
(striped blocks). Ordinate: number of cases, abscissa: “weight difference” in 100

grams.

Generally in experiments with the intact arm hanging verti-
cally, the elimination of the arm-weight causes an underestimation
of the weight to be lifted, so that the weight difference correspond-
ing to the difference threshold in the two unlike postures is rather
big and positive. The additional weight of 2,900 g was the biggest
one required to make the weight lifted in the tilting position just
noticeably hea'vder than the 3,000 g standard weight lifted in the
erect posture. Such a large difference occurred, however, rather
seldom, amounting only in five cases to 2,200 — 2,900g. (It may be
noticed, that in all but one case the difference threshold weight
determined in the erect posture (control experiments) was never-
theless of the usual 100 — 200 g magnitude.) Fig. 4 shows the
distribution of the weight differences corresponding to the differ-
ence threshold. The weight difference shown in Fig. 4 (later on

PERCEPTXOX OP WEIGHT AND PHENOMENAL REGRESSION. 123

simply called "weiglit difference”) indicates hoio much greater the
additional weight must be in the tilting exferiments than in the con-
trol experiments to make the variable weight phenoynenally heavier
than the standard toeight. (Thus:

Diff. threshold — diff. threshold = "weight difference”).

main exp. control exp.

The values thus show the effect of the tilting posture on the dif-
ference threshold independently from the possible difference in
the magnitude of this treshold in the control experiments. As may
he seen from Fig. 4 the difference in the experiments with the
intact arm is usually 400 — 600 g (in 22 cases), seldom less (in 8
cases), but very often 700 — 2,800 g (in 24 cases). As the difference
threshold weight is on an average 79 i 97 g, only differences
greater than 300 — 400 g may be considered significant. According
to the tilting experiments ivith the intact arm, the arm-weight
seems thus to influence the weight perception in 35 cases out of 53
(one case is omitted owing to the abnormally high difference
threshold weight of 900 g in control experiments). In 19 cases no
certain influence was observed. If the results of these experiments
are compared with similar ones performed with the stump arm, it
can be seen, that the influence is real and not for example a laby-
rinthine effect. In case of experiments with a forearm stump the
maximum of the "weight difference” has, as can he seen from Fig. 4,
shifted only slightly to the left (smaller values), compared with the
maximum of "weight difference” in the experiments with the
intact arm. In the arm stump experiments, however, the greatest
number of cases shoio a difference of -^0 (these cases are summed
up in one column), the maximum being definitely more to the left
than in the other experimental groups.

Having established the fact, that the weight of the arm obviously
influences the perception of weight in some cases, the question
has to be settled, why this influence is lacking in as many cases as
shown by the experiments. In the first hand it should be decided,
whether there is a question of a central conceptual constancy
phenomenon or whether the receptual basis of the weight sensation
in these persons, and regardless of the rather heavy standard
weight, might be the intensity of skin pressure, in which case the
arm-weight would not of course be expected to influence the per-
ception of external weights. The question is not quite easy to settle,
but the following discussion might give some indication of what
sort of relation there is between the peripheral on central correlates

124

EEVA JALAVISTO.

of the "Weight perception. If we examine the statements of the
subjects concerning the phenomenal weight of the amputated
arm compared with that of the intact arm (whether the amputated
arm is heavier, equal to or lighter than the intact arm) these
judgements should vary at random if the perceptual correlate is
always peripheral and the receptory mechanism tactual. If, on
the contrary, there is no difference in weight perception in the tilt-
ing and the erect posture, and this would depend on a central
transformatory constancy, the distribution of the statements
concerning the phenomenal weight would be quite definite: the
statement "the stump arm lighter than the intaet arm” should
correlate with a lack of difference in weight sensation between
tilting and erect posture and "vice versa. A special consideration
should be given such cases in which the phantom arm feels lighter,
but there is no "weight difference”. The least complicated expla-
nation in this case would namely be, that the resective basis is
constantly peripheral, either proprioceptive, when the weight
of the own arm is judged, or tactual, when the sensation of weight
arises in connection with the lifting of an external weight resting
on the skin. In this latter case, the muscle tension developed would
of course in no way interfere with the sensation of weight. Tig. 5
shows the distribution of the statements "lighter'' and "of equal
weight” plotted against "weight difference” in experiments "with
the intact arm. The ordinate represents the number of cases, the
abscissa the "weight difference”. As may be seen, there is no
quite firm relationship between the phenomenal weight of the
amputated arm and the "weight difference” found in the tilting
experiments. A clear tendency can, however, be observed that the
statements "of equal weight” are more numerous in such cases in
which the “weight difference” does not significantly differ from 0.
When there is a weight difference, however small and on the limit
of significance, i.e. 500 — 600 g, the number of statements "of equal
weight” and "lighter” is the same. A^^len on the other hand the
“weight difference” is great, the amputated arm is more often
lighter than the intact arm. If the cases "nith just significant
“weight difference” are marked "with 0, the cases without any
“weight difference” with — 1 and those in which the "weight dif-
ference” is obviously significant with -)- 1, the correlation between
the series can be calculated with the usual moment product
method. The correlation coefficient thus obtained is 0.42 0.15

(= 2.8 or) and = 5.6. This means that the cases with no "weight

PERCEPTION OP WEIOnT AND PHENOMENAL REGRESSION. 125

difference” and tiiose with a marhei "weight dijjffre'nce" , dijfer
■signi{ica 7 i(Iy according to the phenomenal weight of the amputated
arm, because the probability that such a distribution might occur
by chance is less than 2 per cent. In three cases only the "weight
sensation has a purely peripheral correlate as the judgement
"lighter” corresponds in the main weight experiments to a lack
of "weight difference” (black columns to the left from the stripled
line in Kg. 5). This correlate is tactual (skin pressure) in these
experiments in spite of the rather heavy standard weight.

liliii I i] 1 1

iO +l'2 df 1*56 *7-8 *Bfl 1516 18 13f0 'Ofl m-yffixIOOg.

weight difference

Fig. C. Distribution of the "weight difforcnco” in cases in wliioh tlio amputated
and the intact arm are judged of equal wciglit (white blocks) and in cases in which
the stump is phonoracnnllp lighter than tho intact arm (blaek blocks). Ordinate:
number of cases, abscissa: “weight difforonco” in 100 grams. To tho loft from tho
stripled line tho "weight difference” is not significant, (Stripled line: border lino
of significance of tlie "weiglit differonco”.)

IVhen trying to find out why the correlation between the
judgements "lighter” and “of equal weight” and the “weight
difference” is not more firm, the possibility must bo kept in mind,
that the state of irritation in the arm stump, which sometimes
makes an amputated arm phenomenally heavier than tho intact
arm, may act in a way that the amputated arm which otherwise
might be lighter becomes phenomenally of equal weight. It should
also be noticed, that it is not always easy for the patient to judge
the weight of his arms; sometimes the phantom arm is otherwise
of equal weight, but if his attention is concentrated upon it and

126

EEVA JAIiAVlSTO.

especially if he swings his stump at the same time, the phantom
arm begins to feel lighter than the intact arm. It may thus be
possible, that the attitude of the different subjects when answering
the question is not the same. On the other hand the real weight
difference between the amputated arm and the healthy one is
objectively smaller in the forearm amputated cases than in the
arm amputated cases and it might be supposed that in the fore-
arm amputated subjects this weight difference might be below the
difference threshold of weight perception. This may be of no great
significance, however, as in both groups of amputated, the cases^
in which the weight of both arms is judged "equal” and the cases,^
in which the amputated arm is judged "lighter”, are equally
numerous.

Discussion.

The results of my experiments differ partly from those of
earlier investigations and partly they are consistent with them.
Firstly, the arm weight seems very often to influence the weight
sensation, when the weight to be lifted is great, contrary to what
is suggested by Katz (1920), in his experiments on amputated
subjects and contrary to the results of Fischel (1926) and Pantzel
(1925) in their experiments on normal subjects. It is very probable,
that this difference depends upon the fact, that in Katz’ and
Fischel’s experiments the indications for weight perception were
impulses derived from the skin (in consequence of the small stan-
dard weight employed) and not from the muscles, as supposed by
Katz. It is true, that in some cases in experiments with amputated
subjects that had Sauerbruch canalized muscles Katz could show
that the receptory basis really was proprioceptive, i.e. the weight
corresponding to phenomenal equality was, when lifted with the
canalized muscle, six times as heavy as a weight lifted with the
intact arm. On this ground Katz supposed that the perceptual
basis would be the same even in different conditions and in differ-
ent subjects. It is obvious, however, that the result cannot be
generalized. Katz’ own experiments show, that in other experi-
mental conditions and with other subjects the indications for weight
perception cannot have been proprioceptive. Difference thresholds
determined in experiments with the canalized muscle and those
with the weight applicated on the forearm did e.g. not differ much
from each other. The muscle tension corresponding to the dif-

PERCEPTION OF IVEIGUT AND PHENOMENAL REGRESSION. 127

ference threshold would, however, in the former case have been six
times greater than the muscle tension corresponding to the dif-
ference threshold, when the weight was lifted wnth the healthy
arm. The only explanation of this controverse might be that in
experiments in which the phenomenal equality is determined
(topological experiments by Eenqvist-Reenpaa (1936)), the per-
ceptual cues might be the tension in the muscles, and in the ex-
periments mth determination of the difference threshold as cues
might serve the impulses caused by the pressure exerted upon the
skin. Considering the perceptual situation in both these cases this
dual behaviour is quite comprehensible; the weight sensations may
be qualitatively quite different owing to the altered sensibility of
the canalization in experiments in which the weights to be com-
pared are lifted alternatively by the intact arm and the canalized
muscle. In this case it would be more adequate to rely upon
the muscular tension, which most probably is a qualitatively uni-
form indication of weight, than upon the qualitatively differing
cues of the tactual sphere.

In the case of determination of the difference threshold, things
are different; Both the standard stimulus weight and the variable
weight are lifted in the same 'way, the sensations of weight being
qualitatively similar in the two weight comparisons. Now there
is no reason to abandon as indications of weight the tactual cues,
which mostly are more accurate than the proprioceptive ones.
The difference threshold weight consequently corresponds to the
pressure exerted upon the skin and not to the tension developed in
the muscles as in the case of the topological experiments.

This shows how necessary it is to consider the receptive
mechanism responsible for the weight sensation in every special
experimental situation.

On the other hand my experiments gave evidence that in some
subjects the weight perception based upon muscular impulses may
he, so to say, controlled by a conceptual character which is demon-
strated by the fact that the arm weight does not, in such cases,
influence the perception of a lifted (external) weight. This behavior
seems to be contrary to Katz', Fischel’s and Pantzel's views,
more uncommon than the weight sensation corresponding solely
to the tension developed in the muscles.

The question arises: on what physiological process is the central
invariance-phenomenon based? Katz refers to the view suggested
by a. E. Muller (Muller and Schumann, 1889) that the weight

128

EEVA JALAVISXO.

perception might depend, on the motor impulse that effects the
lifting of the weight. Thus a weight lifted with a greater impulse
might be lifted with a higher speed and consequently judged lighter.
According to Katz the central nervous system might be so well
adapted to the weight of the own arm, that the speed, with which
the weight is lifted might, in all situations, be independent of the
rotatory moment of the arm against gravity and the judgement
of weight not invalidated by differences in it. Later on several
investigators have shown that the judgement of weight is not
influenced by the indirect evaluation of speed but is based on
the direct perception of muscle tension or pressure on the skin
(Eenqvist et al., for references see Jaxavisto 1935). Thus the
adjustment of the motor impulses cannot explain the phenomenal
regression phenomenon. As a matter of fact the question is difficult
to solve experimentally as the main feature of the constancy phe-
nomenon is precisely that, that a peripheral receptive correlate
is lacking in a strict sense, and the only correlation to be found
is the one between the weight sensation and a conceptual character
of a “real” weight. In these experiments on lifted weights just as
in other constancy (visual) phenomena (Brtinsvik 1934, Thotj-
LESS 1931 — 32) the phenomenal character corresponds very often
to an intermediate value between the stimulus and the conceptual
character of the thing to be perceived.

Summary.

1. The purpose of this paper was to investigate whether the
weight of the subject’s own arm influences the perception of an
external weight to be lifted, when the receptive mechanism is
proprioceptive (as is almost the case e.g. when lifting heavy
weights).

2. The receptive mechanism of weight perception was briefly
discussed.

3. The experiments were performed on 54 war invalids ampu-
tated on the upper limb. The experimental arrangement was as
follows: The subject had to compare two weights, the one of which
(a standard of 3,000 g) was lifted by the arm (the intact arm, or
the stump) in an abduced position, the subject standing erect and
the armweight acting against gravity, the other one (variable
weight) lifted with the subject in a tilting posture, the arm hanging

PERCEPTION OP WEIQUT AND PHENOMENAL UEGRESStON. 129

vertically down and tlie tissues tlius passively supporting tlio arm
weight. The weight to he added to the variable weight in order to
make the two weights phenomenally equal, showed great individual
differences.

4. In experiments with the intact arm, in 19 cases, the additional
weight was at the most 300 — 400 g greater than a weight that
corresponds to the difference threshold determined the standard
weight and the variable weight both lifted in the erect posture
and the arm-weight acting against gravity. The difference is not
statistically significant. In 11 cases it was 500 — 600 g greater,
i. e. on the limit of significance, and finally in 24 cases 700 — 2,800 g
greater, which difference is surely significant. In experiments
with forearm stumfs the corresponding figures were 16 : 4 : 8, and
in experiments with artn shnnfs 19 : 1 : 2. It is obvious, that when
experimenting with the whole arm, iltc arm-ioeight influences the
'perception of an external weight in most subjects, in experiments on
forearm stumps the influence is imich less aiul in experiments on arm
slumps hardly observable, quite in accordance with the fact, that the
weight of the forearm stump is of course a little less, that of the arm
stump considerably less than the weight of the intact arm.

5. Those cases in which (in experiments with the intact arm)
the perception of a lifted weight was not affected by the arm
weight (19 cases) proved, that the proprioceptive weight perception
shoics phenomenal regression to a conceptual thing character analogous
to the well-known constancy phenomena in the visual field. In a few
cases, however, there was reason to believe, that the receptive
mechanism was tactual (pressure exerted upon the skin). In these
cases the weight perception is of course independent of the arm
weight.

Financial aid for this research lias been granted by the Kokde-
LIN foundation.

I hereby wish to express my gratitude to the Head of the Hed
Cross Invalid Hospital, Dr. Eehnberg for his permission to
carry out the investigation on patients at the Hospital. It is a
pleasure to record my indeptedness to the nurses and social case
workers for their land assistance in selecting and sending the
patients over to the Physiology Institute. I am also grateful
to Miss LrrsA Laine, for her valuable help and thoroughness in
carrying out a part of the experiments.

D — tGOSIS. Ada phyx. Sciuidinav. Vol.ll.

130

EEVA JALAVISTO.

References.

Aieila, K,, and E. Jalavisto, Skand. Arch. Physiol. 1939. 82. 136.
Alleks, E., Mschr. Psychiat. Neurol. 1909. 26. 116.

Beunswik, E., Wahmehmung und Gegenstandswelt. Leipzig undWien
1934.

Fischel, H., Z. Psychol. Physiol. Sinnesorg. 1926. 98. 342.

Holmes, G., Lancet 1922. 100. 1177. 1231.

Holmes, G., Ibidem 1922. 103. 59. 111.

Jalavisto, E., Acta Soc. Med. Fennic. "Duodecim” Ser. A. 1935.
X7III.

Jalavisto, E., A. Kalin and L. Parvela, Skand. Arch. Physiol. 1938.
79. 63.

Jalavisto, E., L. Liukkonen, Y. Eeenpaa and A. Wilska, Ibidem.
1938. ,79. 39.

Katz, D., Z. Psychol. Physiol. Sinnesorg. 1920. 85. 83.

Katz, D., and W. Stephenson, Brit. J. Psychol. 1937. 28. 190.

Mann, L., Zbl. ges. Neurol. Psychiat. 1912. 31. 1356.

Martin, L. J., and G. E. Muller, Eef. Z. Psychol. Physiol. Sinnesorg.
1900. 24. 146.

Matthaei, E., Pfliig. Arch. ges. Physiol. 1924. 202. 88.

Merkel, J., Philos. St. (Wundt) 1889. 5. 253.

Muller, G. E. and F. Schumann, Pfliig. Arch. ges. Physiol. 1889. 45.
37.

Pantzel, a., Dtsch. Z. Nervenheilk. 1925. 87. 161.
Eenqvist-EeenpaX, Y,, Allgemeine Sinnesphysiologie. Wien. 1936.
Eenqvist, Y., Z. Biol. 1927. 85. 391.

Eenqvist, Y., K. V. Bagh and E. Elmgren, Skand. Arch. Physiol.
1932. 63. 285.

Thouless, E. H., Brit. J. Psychol. 1931. 21. 339.

Thouless, E. H., Ibidem 1932. 22. 216.

Wangel, E., E. Elmgren, K. v. Bagh and Y. Eenqvist, Skand. Arch.
Physiol. 1931. 63. 133.

Weber, E. H., Annotationes anatomicae et physiologicae de pulsu,
auditu, tactu etc. Programmata collecta. Lipsiae 1834 et 1851.

From Univorsitetets Biokeniiskc Institut, Copenhogcn.

On tlie Synthesis of Creatine in the
Animal Body.

GUNNAB STBENSHOLT,

Received If) Noveinlicr lOR').

In recent years tlie problems connected with the s/ntliesis of
creatine in the animal body have occupied the attention of a
number of biochemists. As a result of the persistent efforts of
these workers it can be said to-day, with a rca.sonably high degree
of probability, that creatine is formed in the animal body by
raethylation of gunnidino acetic acid, the methyl groups being
furnished by suitable methyl donors among which in all pro-
bability methionine plays a very prominent part. The present
unriter recently had occasion to present a short review of these
developments together with some new experiments which strongly
corroborate the assumption just outlined of the mechanism under-
lying the biosynthesis of creatine.

It is a serious defect of much of the work done on creatine
— and not only work connected with the biosynthesis of the
substance — that the methods ordinarily employed for the quan-
titative determination of the compound are rather unspecific.
Tin's criticism also applies to the investigations of the present
writer (Steensholt (1945)). It is, therefore, a matter of consider-
able interest to check previous rvork in this field by new and
more specific methods whenever they become available.

Probably the most widely used method in studies on creatine
is that of Foltn (or one of its modifications), ■which is based on
the so called reaction of Jaffe, ?. e., the fact that a red colour

132

QUNNAU STEENSUOLT.

develops when an alkaline solution of creatinine is treated with
picric acid. This method was also used in the work of the present
writer. Unfortunately, however, also other substances than crea-
tinine, for instance glucose, react %vith picric acid, and this may
seriously disturb quantitative determinations by this method.
It is therefore of great interest that Benedict and Behre (1936)
and Langley and Evans (1936) found in 3,5-dimtro-benzoic
acid a reagent for creatinine Avith a fairly high degree of speci-
ficity. The Avorkers mentioned found that the colour reaction of
creatinine and 3,5-dinitro-bcnzoic acid is not disturbed by the
presence of the folloAving substances;

Glucose Guanidine

Creatine Methyl guanidine

Glycine Dimethyl guanidine

Guanidine acetic acid Eructose

Arginine Cystine

Acetone and acetoacetic acid Avcre found to interfere Avith the
reaction. HoAvever, in Avork on muscles or muscle extracts these
substances are hardly present in amounts sufficiently large to
cause any difficulties.

As compared to the old method o'" Eolin a method bas9d on
the new reaction Avould evidently be one of A'cry considerable
specificity. HoAvever, before a satisfactory procedure can be dc-
A’^$loped, tAvo main obstacles haA'c to be overcome, namely:

1) the violet-red colour Avhich develops by the interaction be-
tween 3,5-dinitro-benzoic acid and creatinine in alkaline solu-
tion, is not A’-ery stable;

2) the 3,5-dinitro-benzoate solution has a rather strong colour
of its OAvn, A’^ery similar to that Avhich dcA’^elops in the presence of
creatinine.

These difficulties, Avhich necessitate, special precautions, appear
to have been successfully OA’’crcome by Lehnartz (1941) in a
paper, in Avhich he describes a method for the quantitative de-
termination of creatine in muscle, based on the principles out-
lined above. The details of his procedure Avill be described beloAV.
Referring to the above remarks on the defects of the Eolin method,
it Avas considered adAusable to check onr previous results (Steens-
HOLT (1945)) by means of the ncAV method developed by Lehnartz.
The present note is intended to gL’^e a report on the results of this
work.

cnEATINE IN THE ANIJUL BODY.

i3r>

Experimental Part.

Biological material. Our biological material consisted throughout;
of muscle and liver tissue from, rats. The animals -were usually from
4 to 8 months old, and wore kept on a diet believed to be sufficient
in all respects. They were killed by decapitation and the organs re-
moved immediately after death. The ti.ssue was placed on a watch
glass and by means of a jiair of bent scissors it was very finely divided
into a homogeneous mass, Avln’ch could be conveniently handled and
weighed. In the e.vperiments to be described below no difference was
found between tissues from male or female rats.

Fig. 1. pll dependence of the metliylation of etlinnol nmiiie. Abscissa: plf;
ordinate: relative increase (in per cent.) of choline in flask A.

Experimental arrangement and method of analgsis. A typical e.vperi-
ment ran as follows: 0.:! g muscle tissue (from the bind legs of a rat)
were suspended in 4 ml ])lio.spbatc buffer (pH around 7.0), to wbicb
had been added 9 mg guanidine acetic acid and 50 mg methionine.
This mi.\turc was incubated at 37° C for IG hours. At the end of this
period 4 ml 20 per cent trichloracetic acid were added for deproteina-
tion, and in addition 1 ml 10 jicr cent h 3 'drochloric acid. The mi.xturc
was centrifuged after 1 hour, aud 5 ml of the supernatant liquid were
autoclaved at 130° C for 30 minutes.

For the now following determination of creatinine the following
reagents were used:

1. 6 per cent dinitrobenzoate solution, prepared in the following
way: 30 g 3.5-dinitro-benzoic acid were suspended in 425 ml water under
mechanical stirring. 75 ml 10 per cent NuiCOa were added and the
stirring continued for 30 minutes until almost complete solution of
the dinitro-benzoic acid. After filtration the solution was clear aud of
a faint yellow colour. The dinitro-benzoic acid was synthesised by the

134

GUNNAR STBBNSHOLT.

writer by standard methods and purified by repeated recrystallisa-
tions from alcohol. A sample of commercial dinitro-benzoic acid was
also available. After purification the two specimens behaved identic-
ally throughout the work.

2. 2.0 n NaOH.

3. 20 per cent sodium acetate solution.

4. Methyl red indicator.

After autoclaving the liquid was allowed to cool down. Some methyl
red indicator was added, followed by sufficient sodium hydroxide to
ensure exact neutralization. Water was then added to make the total
volume of the liquid 11 ml. Then were added 10 ml dinitro-benzoate
solution. 10 ml sodium acetate solution and finally 1 ml sodium hydrox-
ide. After shaking the flasks were left standing for 5 minutes. At the
end of this period their contents were brought quantitatively over
into 50 ml flasks, which were subsequently filled up to the mark with
water. The final colorimetric measurement was carried out in a Pul-
frich photometer (using filter S 57). As comparison solution a solution
was prepared in exactly the same way as above from dinitro-benzoate,
sodium acetate and alkali.

A blank was carried out simultaneously on a reaction micture con-
sisting of 0.3 g muscle tissue and 9 mg guanidine acetic acid in 4 ml
phosphate buffer. The working up of the mixture was, of course, exactly
as described above.

On subjecting these two reaction mixtures to colorimetric measure-
ment, we obtained in the first case the photometer reading 3.oo as
the mean of five consecutive readings, and, similarly, in the second
case the photometer reading 2.70. This indicates a very considerable
increase in total creatinine in our reaction mixture as compared to the
blank.

This procedure which was illustrated above by an example
chosen at random from the laboratory journal, was systematically
varied with respect to the relative amounts of tissue, guanidine
acetic acid and methionine, and also with respect to the duration
of the incubation, the amount of threechloro-acetic acid and hydro-
chloric acid added, and the time of autoclaving. The results were
qualitatively the same in all cases and further numerical details
will therefore be left out.

By carrying out experiments such as the one just described
in Mcllvaine’s phosphate = citrate buffer at varying hydrogen ion
concentrations a pH-activity curve for the process was obtained.
In the accompanying diagram we have plotted the relative in-
crease in total creatinine expressed in per cent as function of
pH. The main feature of the diagram is of course the fairly broad
optimum somewhat above the neutral point. Several experiments
of this type were carried out, but further numerical details are

CKEATINK IN THE ANIJIAL HODY.

135

left out, since it is felt that the diagram given is sufficiently
illustrative of the general nature of the results.

AVork on liver tissue gave similar result.

(.loiiniientury.

On comparing the results presented above to those obtained
by Foltn’s method and reported in a previous paper (Steen.s-
BOLT (1945)) it is seen that both .sets of measurements agree in
a very satisfactory way. It may therefore l)e concluded that the
results of the present paper strongly corroborate those previously
given.

The MTiter is glad to express his best thanks to Professor licK
for his generous hospitality and support.

Suniinnry.

The paper describes the results obtained on applying the method
of BENEDICT-BEBRE-LANGLEY-EVAKS-LEHNAim for the (piaU-
titative determination of creatinine to the problem of the racthyla-
tion of guanidine acetic acid in the animal body. It is found that
muscle and liver tissue from the rat arc capable of catalysing the
methylation of guanidine acetic acid to creatine, the methyl
groups being furnished by methionine. The results thus corrobo-
rate those of a previous paper.

Hoforoncos.

Benedict, St. B., and J. A. Behre, .1. biol. Chem. 1930. 114. 515.
Langley, AV. J., and M. Evans, Ibidem 193G. 115. 333.
Lehnartz, E., Hoi'pe-Seyl. Z. 1941. 271. 265.

Steensholt, G., Acta physiol, scand. (In the press).

From Universitetets Biokemiske Institut, Copenhagen.

On Methylation Processes in Etiolated
>Vlieat Grerms.

GUNNAR STEENSHOLT.

Received 15 November 1945.

For a number of years tlie methylation processes in living cells
have aroused the keen interest of a number of biochemists. Prob-
ably the best known examples of processes of this kind are

1) the methylation of guanidine acetic acid to creatine; and

2) the methylation of ethanol amine to choline.

In both reactions methionine can function as methyl donator.

Most of the work in this field has been done on animal tissue
(for a summary of much of the work see Guggenheim (1940) and
Werle (1943); a very brief survey of recent contributions is con-
tained in Steensholt (1945)). Great interest was therefore aroused
Avhen Barrenscheen and Pany (1942) reported some experiments
which proved that etiolated wheat germs were able to transform
guanidine acetic acid into creatine. The formation of creatine
was found to be increased by a factor of 6 or even 8 on addition
of methionine to the reaction mixtures. It was further found by
Barrenscheen and Valy (1943) that etiolated wheat germs
transform glycine into betaine, methionine again acting as methyl
donator.

These results have suggested to the present writer the desira-
bility of investigating whether etiolated wheat germs are also
able to methylate ethanol amine to choline in analogy Avith results
recently obtained for animal tissue (Steensholt (1945)). The
present note is a report on the results.

M ETHYLATION pnOCESSES IN ETIOLATED WHEAT GERMS. 137

Experimentnl Part.

Biological material. The wlicnt used was of tliree different kinds:

1) Jubilee Avbeat;

2) Svalof Skandia wbcat;

3) Eecord wheat.

The etiolation wa.s carried out in the usual way by placing the germs
on wet filter paper in the dark. The etiolated wlicat germs Avere used
after G — 7 days. They Avero first cut finely into pieces by means of a
pair of scissons, and afterwards ground in a mortar. The resulting homo-
geneous mass could be conA-cnicntly handled and Avcighed.

Determination oj choline. We decided to apjily a colorimetric method,
since this is probably the more convenient procedure for scries deter-
minations. For this puiposc two methods arc preferred today;

1) Determination as choline iodide (Roman (1930));

2) determination as choline rcineckate (Beattie (1930)).

The method of precipitating choline ns a rcineckc salt is one of con-
siderable specificity, and is probably the procedure most AA-idoly used
to-da)’. The determination Avas usually carried out by measuring colori-
metrically the reddish colour imparted to acetone by choline rcineckate.
HoAvcA’cr, only relatively concentrated solutions are sufficiently strongly
coloured to yield accurate determinat ions. It may therefore be regarded
as a considerable progress AA'hen Ro.s.si, Mauenzi and Lono (1912)
studied photometrically the method for determining the chromate ion
(Cro^ by a procedure based on the reaction of Cazkneua'E (1900), and
found that it could be used for the determination of chromium in
rcincckatcs. On this basis Mauenzi and Cardini (1942) doA’cloped a
ncAA' method for the determination of choline. Arliich is claimed to be
considerable more sensitiA'e than those previously described in the
literature. Wo therefore decided to apply the method of Mauenzi and
Cardini in the present piece of iivA'estigation.

The technical details of the procedure are as folloAA’s:

Reagentn.

Saturated solution of ammonium rcineckate in distilled Avater, pre-
pared immediately before use. The concentration of the solution is
approximately 4 per cent.

96 per cent alcohol.

100 per cent acetone.

60 per cent acetone.

10 per cent NaOH.

10 per cent (by Amlumc) sulplmric acid.

0.2 per cent diphenyl carbazide in 96 per cent alcohol. This solution
first acquires a faint rosy colour, which deepens after a fcAv days. The
solution can be used, nevertheless. The present AA'ritor AAmrkcd Avith solu-
tions which Averc never alloAved to become more than 4 to 5 days old.

The amount of choline to be determined A-aries betAA'ceu 15 v and
100 y.

138

GUKNAR STEBNSEIOLT.

The volume of the sample to be analysed may range from 1 to 3 ml.
The sample is placed in a centrifuge tube with slender end and an equal
volume of a saturated aqueous solution of ammonium reineckate is
added. The tubes are then cooled in ice water for at least 20 minutes.
Longer cooling is, however, superfluous. The mixture is now centrifuged
for 4 minutes (at 3,000 E.P.M.), and the supernatant liquid is after-
wards removed as completely as possible without loss of precipitate
by means of a fine tube provided ivith a suction bulb. The precipitate
is washed with 0. 5 ml ice cold alcohol two or three times. This operation
must be carried out with care lest some of the precipitate be lost.
The tubes are now again chilled for a few minutes in ice water and then
centrifuged. The washings are repeated as described above. The super-
natant liquid is now usually colourless, but sometimes a third washing
has proved necessary.

The precipitate is now dissolved in about 1 ml of acetone and the
solution is transferred to an ordinary test tube. The centrifuge tube is
washed carefully 2 or 3 times with 1 ml 60 per cent acetone each time,
and the collected washings are added to the solution in the test tube.
We then add: 2 ml water, 0.2 ml sodium hydroxide and 0. i ml perhydrol,
for each 50 y choline in the sample. Thus prepared the tube is placed
in a boiUng water bath. In the start the heating must be conducted
carefully due to the rapid evaporation of acetone to begin with. After
most of the acetone is evaporated the tubes are kept in the bath
for 20 to 30 minutes, and finally they are heated over a naked flame
for a few seconds in order to ensure complete elimination of the
perhydrol.

During the heating the liquid acquires a yellow colour.

After the oxidation of the chromium has been completed the tubes
are cooled and the contents diluted with 3 or 4 ml of water. 2 ml sulphuric
acid are added together with sufficient diphenyl carbazide solution to
give a final concentration of 8 per cent. The reaction mixture is finally
diluted to an appropriate volume in a suitable measuring flask, in our
work to 25 ml. The photometric measurements were carried out with the
Pulfrich photometer, using filter S-53. The comparison tube contained
a blank consisting of 2 ml sulphuric acid and 2 ml diphenyl carbazide
solution made up to a final volume of 25 ml.

A calibration curve is conveniently used.

Substrates. The methionine was a Hoffman-la Eoche product. The
ethanol amine was synthesised by the writer according to Knorr (1897).
The method of Knorr consists in leading a stream of ethylene oxide
through a concentrated aqueous ammonia solution and subsequently
fractionating the reaction mixture. This method was found to work
very satisfactorily for the purpose of the present investigation.

A typical experiment was carried out as follows:

In a small flask A were placed

1 g plant tissue

0.2 00 ml ethanol amine

100 mg methionine

6 ml phosphate buff r ^pH ~ 7.o).

METHYLiTION PROCESSES IN ETIOLATED WHEAT GERMS. 139

A similar flask B contained exactly the same amounts of tissue and
reagents but no methionine. Both flasks were incubated at 37° C for 12
hours. At the end of this period 4 ml 20 per cent trichloroacetic acid
were added in order to remove proteins. After centrifugation 2 ml of
the supernatant liquid were removed for choline analysis according
to the method of ]\Iarenzi and Cardini. Double analyses were always
carried out. No difference in choline content between the two flasks
could ho found.

The relative amounts of tissue and reagents were systematical^
varied, as was also the time of incubation (up to 48 hours). Experi-
ments were also carried out at 24°0. Qualitative!}’ the results were
the same in all cases; no difference in choline content between
flask A and flask B could be found. The three types of wheat with
which our -work was carried out, behaved identically.

Barrenscheen and Pany (1942) in their work on creatine
synthesis in etiolated wheat germs found it necessary to oxygenate
the reaction mixtures. It might therefore be supposed than oxygena-
tion of the reaction mixture is required also for choline synthesis.
E.xperiments to test this point Avere consequently carried out, but
the results were identical to those just reported.

According to some previous experiments of the present writer,
the methylation of ethanol amine to choline by animal tissue Avith
methionine as methyl donator seemed to go somewhat better in
MoIlvaine’s phosphate-citrate buffer than in other buffers. Ac-
cordingly this buffer Avas tried also for the purpose of the present
work, but the results remained unchanged.

Comments.

According to the experiments reported above etiolated Avheat
germs are unable to catalyse the methylation of ethanol amine
to choline, A\dth methionine as methyl donator. This is rather
interesting in vicAv of the results of Barrenscheen and Pany
(1912) already referred to, and shoAvs that an enzyme or enzyme
complex capable of catalysing the transfer of methyl groups from
methionine to guanidine acetic acid is ineffective AA’hen ethanol
amine is the acceptor. Eor animal tissue the corresponding specific-
ity problem is still unsolved.

140

GUNNAR STEENSHOLT.

Summary.

It is found that etiolated -wheat germs are unable to catalyse the
methylation of ethanol amine to choline, with methionine as
methyl donator. The meaning of this is briefly discussed.

The -writer is glad to express his best thanks to Prof. Ege for his
generous support and hospitalitj’-.

Heferenees.

Baerenscheen, H. K., and J. Pany, Biochem. Z. 1942. 310. 344.
Baeeenscheen, H. K., and S. Valy, Hoppe-Seyl. Z. 1943. 277. 97.
Beattie, F. J. R., Biochem. J. 1936. 30. 1554.

Cazeneuve, a.. Analyst 1900. 25. 331.

Guggenheim, E. A., Die biogenen Amine, Basel — ^New York 1940.
K^norr, L., Ber. dtsch. chem. Ges. 1897. 30. 909.

Marenzi, a. D., and C. E. Cardini, J. biol. Chem. 1943. 147. 363.
Roman, W., Biochem. Z. 1930. 219. 218.

Rossi, L., A. D. Marenzi and R. Lobo, An. farm. y. bioquim. 1942.
13. 1.

Steensholt, G., Acta physiol, scand. (In the press).

"Werle, E., Die Chemie 1943. 56. 143.

From the Depnrtmcnt of Chemistry, the JRoynl Veterinary Institute,
Stockholm, and tlic Kristineberg Zoological Station, Sweden.

Choline Esterases in some Marine
Invertehrales.

KLAS-BERTIL AUGUSTINSSON.

Bfccived L’t November 1915.

The presence of choline esterase (CIiE) is necessary for any
function of acetylcliolino (AC1») and is, therefore, of great signifi-
cance. ACh jnetaholisni is, in itself, “intrinsically connected with
the electrical changes dtiring nerve, actvity, occurring every-
where of the neuronal surface” (Bublock and •Nachmansohx
1942). Broni a comparative ])oint of victv, it i.s of interest to
determine the ChE activity of different species of all animal
groups; we want to know whether the ACh mechanism has been
evolved parallel to the differentiated nervous systems.

This is the physiological point of the ])rol)lem, which lias al-
ready been investigated by several authors. A summary of these
investigations is given below and in Table 1.

In the Protozoa neither ChE nor ACh are to be found. As a
rule, Coelenterata do not show any ChE activity. In worms,
considerable quantities of the ACh hydrolysing enzyme are
present. The blood, however, seems to have no ChE activity.
Among the Crustacea, particularly the lobster and the crayfish
have been investigated, but other species have also been examined.
The enzyme is to be found in muscle and nervous system, but is
lacking in the blood. This is also the case with the spiders and the
insects. Considerable amounts of ChE are found in the blood of
the Jlollusca. The activity is the lowest in the mussels (Lamelli-
brancliiata), a little higher in the squids (Cephalopoda), highest
in the snails (Gastropoda). High concentrations of ChE arc found
also in the muscles. ChE is lacking in the purple cyst of Murex
(JuLLiEX 1939; JuLUEN and Bonnet 1941). It is present in the

142

Kr.AS-BERTlL AUGU.'TISSSON.

Table 1.

Protozoa

lIlTROPOLITANSKAJA 1941; BtJLLOCK and NACmiAXSOHIt
1942

Coclcnterata

Sponges

Bacq 1935, 1937 b; Mitropolitanskaja 1941

Hydrozoa

Bacq and Oury 1937; Mitropoutakskaja 1941; Bullock
and Nachmansohn 1942

'Scyphozoa

Bacq 1935; Mitropolitanskaja 1941; Bullock and
Nachmarsohn 1942

Anthozoa

Bacq 1935, 1937 a; Bacq and Nachaiansohr 1937;
Mitropolitarskaja 1941; Bullock and Nachmansorr
1942

Ctenophora

Bacq and Oitry 1937; Bullock and Nachjiarsohn 1942

Worms

Turbellaria

Bacq 1937 a; Bullock and Nachmarsohn 1942

Trematoda

Bacq and Oury 1937

Ccstoda

Aetemow and Lurje 1941

Nematoda

Bacq and Oury' 1937

Polyohaeta

Bacq 1935, 1937 a; Halperr and Corteooiari 1935;
Bacq and Oury' 1937; Bacq and Nachmarsohr 1937;
Kiechert and Schrarreneerger 1942

Sipunculoidca

Bacq 1935 b, 1937 a

Crustacea

Bacq 1935, 1937 a; Koschtojartz 193G; Bacq and Nach-
iiARSOHR 1937; Bacq and Oury 1937; Marray and
Nachjlarsohr 1937; Jullier and Vircent 1938; Ar-
TEMOW and Mitropolitarskaja 1938; Nacmmarsohk
1938, 1939; Nachyunsoiir and Bothenbero 1945

Spiders

Bacq and Oury 1937; Corteooiari and Serfaty 1939

Insects

Bacq 1935; Bacq and Oury 1937; Corteooiari and
Serfaty 1939; Tahmisiar 1941; Mears 1942

Jlollusca

Gastropoda

Bacq 1935, 1937 a; Gautrelet 1935; Halperr and
Corteooiari 1935; Amsior 1935, 1943; Koschtojartz
1936; Bacq and Oury 1937; Vircert and Jullier
1938, 1939; Joluer, Vircert, Bouchet and Vuillet
1938; Jullier 1939, 1941; Jullier and Bokret 1941;
Rezek and Haas 1942

Lamelli-

Bacq 1935; Viroent and Jullier 1938, 1939; Jullier,

branchiata

Vircert, Bouchet and Vuillet 1938

Cephalopoda

Bacq 1935, 1937 a; Bacq and Nachmarsoiir 1937; Jul-
lier, Vircert, Bouchet and Vuillet 1938

Echinoderma

Bacq 1935, 1937 a; Bacq and Nachmarsohr 1937; Bul-
lock and Nachjiarsohr 1942

Tunicata

Bacq 1935, 1937 b

blood, tlie muscles, and tlie nervous system of echinoderms.
Tunicata have no ChE in tbe blood, but tbe muscles bave an
ACh hydrolysing activity.

Tbe author of this paper lias • determined the ChE content of
different invertebrates chiefly from a chemical standpoint.

The last four years of investigations of this enzyme have
indicated that the ChE activity, in many cases, may not be
attributed to only one onzjune; inore types — 2 at least — are
to be found.

CHOLINE ESTERASES IN SOME MARINE INVERTEBRATES. 14?)

In the case of these tivo ChE, wc know with all probability that
the cnzTOie in the erythrocj'tes is a specific choline esterase.
On the other hand, there is no evidence in favour of a specific
choline esterase in blood serum. This was first shown by Vaiii.-
QUiST (1935) in a cataphoretic investigation. Doubt that this
“serum choline esterase” was identical with the acetylcholine
hydrolysing enzyme in the er\i;hToc)d:es was first expressed by
Alles and Hawes (1940, 1941, 1944). This doubt was soon con-
firmed in different quarters. Thus Richter and Croet (1942),
Mendel and Rudney (1943, 1944-, sec also Mendel 1943; Mendel,
Eudney and Strklitz 1944), Zeller and Bissegger (1943),
and Eachm.ansohn and Eothenberg (1945) showed that ChE
in blood scrum and certain tissues was a non-siiecific enzyme.
As regards the blood, definite proof has been established by this
author (Augustinsson 1944, 1945) in a cataphoretic investigation.
JIendel and Rudney have named the ACh hydrolysing enzyme
in serum “pseudo choline esterase”, which has been described as
an “unfortunate designation” (Laurenfels 1943).

All those papers have given rise to the assumption that we have
to deal ivith two distinct choline esterases; one specific, and one
unspecific ChE. The specific enzyme is found in erythrocytes
(in some cases in the plasma; Mendel, Mundell and Rudney
1943), in brain, and other nerve tissues, as well as in the electric
organ of Torpedo (Nacumansohn and Rothenberg 1945).
L.angejiann (1944) has, however, reported that the enzyme in
the anterior pituitary lol,>e has not the same properties ns that
in the posterior lobe (and the erythrocytes). He also found the
erythrocyte type in skeletal muscle (also Nacumansohn and
Rothenberg 1945) and, in certain cases, in the thymus. A mix-
ture of the two types of enzymes is indeed present in most of
the blood sera and many tissues, as well ns in the superior cervical
ganglion (Mendel and Rudney 1943, 1944; ]\Iendel, Mundell
and Rudney 1943). According to Glasson (1944), a mixture is
present in red corpuscles. The difference between the ChE in brain
and serum has also been investigated by Schab-Wotrich (1943).

J^lethods for estimations of the two types of ChE have been
described by Mendell, Mundell and Rudney (1943). These
methods are based on the use of two different choline esters as
substrates. Thus acetyl-^-methylcholine is hydrolysed by the
specific enzyme only, benzoylcholine by the unspecific (pseudo-
ChE). Both types catalyse the hydrolysis of ACh. It is, however.

144

KLAS-BERTIL AUGUSTtNSSON.

extremely doubtful whether such a test really is a demonstration
of the whole point. That this doubt is justifiable, is obvious from
the following investigations, described in this paper and others
to follow. The results are obtained with different species of prim-
itive animals.

Methods.

Immediately after the animals were caught, they were frozen in at
— 20° C. The frozen material was melted and minced. About 5 g.
were taken for further grinding in a mortar with washed sand. The
disintegrated tissue was then taken up with, in most cases, twice as
much bicarbonate-Ringer solution (Rso). The mixture was shaken for
some minutes, centrifuged at a constant speed of 3,000 r. p. m., and the
fluid decanted. Very different time intervals were needed to produce
as clear a fluid as possible. The buffered suspensions were kept in the
refrigerator, and the analyses were made, in most cases, the following
day, in others, after two or three days.

The ChE activity was measured with the manometric method by
Warburg, in the same way as described in a previous paper (Augus-
TiNSSON 1944). In the main compartment of the flask, l.G ml. of the
solution of the substrate was placed; in the side bulb 0.4 ml. of the
enzyme solution. For the dissolution of the substrates the same bicar-
bonate-Ringer solution (Rao) was used, as in the preparation of the
enzyme suspensions. The composition of R 30 was; 100 ml. O.o %
NaCl + 2 ml. I.2 % KOI +2 ml. 1.7 6 % CaCb (cryst.) -f '30 ml.
1.2 0 % NaHCOa. Fresh Ran was used for each experiment, since the
solution deteriorates if kept.

The hydrolysis was carried out in a gas mixture of 95 % Na and 5 %
CO 2 , and at 37.. 0 ° C. Readings were made continuously for 60 minutes
for the period 5 to 65 minutes after addition of the enzyme to the
substrate solution. The thermobarometer was filled with 2.0 ml. of the
substrate solution with the same concentration as that in the reaction
solutions.

Table 2.

Substrate

■

ml.

solution

Total cone.

Vjp = 2.00

a — b

■1

ACh

181.7

0.25 %

l.G

0.20

0.011

MeCh

195.7

0.010

BzCh

243.6

0.008

ACh = acetylcholine ehloride (Hoffsian-La Roche); MeCh = aeetyl-/S-niethyl-
choline chloride (Merck); BiCh == benzoyloholine chloride (Hoffmaxn-La
Roche).'

' The author wishes to acknowledge with thanks the gift of benzoyloholine
chloride by Messrs. Hoffji.ank-La Roche.

CHOLINE ESTEKASES IN SOME MABINE INVERTEBKATES. 145

The total volume of CO 2 evolved during 60 minutes (enzymic + spon-
taneous hydrolysis) is expressed as a [A., the enzymic hydrolysis as b
/i\. (total — spontaneous).

Table 2 shows the amounts and concentrations used of the substrates.
The concentrations of the substrates were constant in all experiments
and were all well above the optimum substrate concentrations in each
case. Tlie amounts of the cnzjmie preparations used are found in
Table. 3.

Results.

In most cases no attempt was made to dissect the animals to
find active tissues. 20 species of marine animals, belonging to 12
different groups of invertebrates, ivere examined.

The activity is expressed as 6/100 mg (Q in Table 3), that is
jA. CO 2 evolved in 60 minutes by 100 mg. tissues.

As regards the hydrolysing effect on ACh, these results con-
firm, on the •whole, the observations made earlier. Strangely
enough, the two 'worms did not show any ChE activity. AVith the
Crustacea no unitary result -was obtained. Pandahts and especially
the muscles of Carcinus presented high activity; the intestines
and the muscles of Eupacfurus, on the other hand, were inactive
(cf. Bacq 1935 a, 1937 a). A Placophora as well as two Gastropoda
gave significant positive values. Large concentrations w'ere found
in Dentalium. The Lamellibranchiata seem to be less ChE active
than the Gastropoda. Positive values were also obtained with a
Brachiopod. Considerable amounts of ChE Avere shown in the
whole animals of Antedon petasus. The other echinoderms were
clearly ChE active. The quotients Avere very loAv for Amphiura,
and among the Holothurioidea Gucumaria Avas about 5 times
more active than Mesothuria.

In most cases, ACh was split at a higher rate than any other of
the two substrates. This result is in accordance A\dth “the main
feature of choline esterase” (Naohmansohk and Bothenberg
1945). In one case ACh only Avas split (Terebratulina). Of the two
other substrates, only Meflh Avas hydrolysed in most cases, for
Patella and the ampulles of the podia of Asterias this effect was
OA’^en more pronounced than Avith ACh. The same amplitude Avas
found in Mya, Amphiura, and Holothurioidea. BzCh, also. Asms
hydrolysed in some cases. The activity aa^s particularly strong
for Tonicella, and the intestines of Asterias, even stronger than
with MeCh. The muscles of Carcinus, also, split BzCh.

According to Mendel and collaborators, the hydrolysing effect

10 i60215. Acta pliys. Scandinav. Vol. 11.

146

KLAS-BERTIL AUGDSTINSHON.

Table 3.

Species

Test susp.

mg.

tiss.

used

ACh

McCh

BzCh j

g. tiss.

Polychaeta

Aphrodite aculeala . . .

1.4

5.0

100

Nereis virens

S.9

17.8

200

Crustacea

Balanus erenalus

2.5

5.0

200

Pandalus montagui . . .

10.0

20.0

200

3.5

Garcinus maenas:
muscles

5.0

10.0

200

298

149

Eupagimts hernhardus;
intestines

2.5

10.0

100

muscles

0.7

2.8

100

Placophora

Tonicella marmoraia . .

1.45

2.9

200

Gastropoda

Patella vulgata

4.95

10.0

200

121

Purpura lapillus

5.0

10.0

200

171

Solenoconchae

Dentalium entails ....

O.G

2.4*

100

134

Lamellibranchiata

Mga arenaria

5.0

10.0

200

Astarte sulcata

i.i

o o

200

Braohiopoda

Terebralulina caput ser-
penlis

2.0

4.0

200

Crinoidea

Antedon petasus

5.0

10.0

200

451

226

139

Asteroidea

Asterias rtibens;
intestines, stomach.

2.9

6.8

200

113

ampulla, podia ....

4.05

8.1

200

Ophiuroidea

Amphiura chiajei ....

3.9

7.8

200

Echinoidea

Psammechinus milia-
ris: intestines

1.45

2.9

200

144

Echinus csculenlus: in-
testines

4.1

8.2

200

169

Holothurioidea

Jfesothuria intestinalis

9.G

19.2

200

■Cucumaria laclea ....

2.5

5,0

200

349

175

299

150

in the material examined must, in most cases, be ascribed to the
specific ChE. In some species, however, there should be a “pseudo
ChE” too, characterized by its power of hydrolysing BzCh. It is,
indeed, of little probability that we in these cases have to deal
with only “two types”, the properties of which are so different in
different material (see below under the Discussion). The two
types should separately be responsible for the hydrolysis of ACh.
Thus we can expect to find about the same proportion between

CHOLINE ESTERASES IN SOME MARINE INVERTEBRATES. 147

the speed of the hydroly.sis of ACh and the sum of speeds, ob-
tained with the two other substrates. A comparison of this kind
is found in Table 4.

Table 4.

Species

QacIi/ (QjIeCh + ^BzCh)

Pandalus

2.9

Carcinus

2.1

Tonicella

1..5

Patella

0.8

Purpura

6.1

Dentalium

2.3

Mya

0.9

Astarte '.

3.5

Terebratulina ....

Antedon

3.2

Asterias

I.l

0.8

AmpTiiura

1.0

Psammcchinus . . .

1.8

Echinus

1.9

Mesolliuria

1.2

Cucumaria

1.2

This table shows that the ratio in question in some cases is very
different in comparison to the others. Particular facts must be
present. The results are hardly to be explained by the acceptance
of only two choline ester hydrolysing enzymes.

In the following discussion I will try to present the facts which
may be the foundation for further studies on this problem.

Discussion.

ACh is an ester, the ester linkage of which in no way differs
from a linkage of the same kind in other esters. Because of this,
there is nothing remarkable in the fact that the enzyme or en-
zymes, hydrolysing ACh, also might split an ordinary ester, e. g.
methyl butyrate. Thus a ChE need not necessarily be specific for
choline esters. This does not mean, however, that serum ChE,
for instance, may be identical with an ordinary esterase. There
are many facts that argue against this, in the first place the in-
hibitory effect of various substances. We have, with all probability,
an unsfedfic ChE in blood serum. Besides the ester linkage,
ACh has a positive ammonium ion which, no doubt, is of great
importance in forming the enzyme-substrate complex. It is to
be assumed that it is at the linkage of this group to the enzyme
that the specific ChE-inhibitor.s react.

148

KLAS-BERTIL ADGUSTIKSSON.

A S'pecific ChE is sucli an enzyme as hydrolyses ACh at a higher
rate than any other esters, but it may fully split them. Thus the
specificity for ChE is relative. The forming of an enzyme-sub-
strate compound is controlled by the substituted ammonium
ion. This positive group may be assumed to be attached to a
negative group in the enzyme molecule (in the apo-enzyme com-
ponent). The velocity of the hydrolysis is controlled by the reac-
tion between the ester linkage of the substrate and the enzyme
(the co-enzyme).

According to the nomenclature of Bergman (1942), ChE should
be a heterospecific enzyme. With the presence of two active centra
in the enz 3 Tne molecule, we may explain the fact that a substrate
concemration above the optimum one has a . depressing effect
on the specific ChE. In this case both centra are attached to the
same substrate molecule, low concentrations, however, to two
molecules (cf. Zeller and Bissegger 1943). This assumption is
confirmed by the fact that strongly positive proteins (Mendel
and Eudney 1944) change the properties of the enzyme in such
way that the characteristic inhibition by an excess of substrates
disappears. This effect of certain proteins is most probably a
neutralisation of the negative group in the enzyme, that group
which combines with the positive N-group. This also explains the
different properties of ChE in different tissues, and suggests the
necessity of using pure enzyme preparations. Compare also the
cataphoretic investigation by the author of this paper (1944,
1945).

The real chemical difference between the two types of ChE,
the specific and the un-specific, is difficult to understand now-
adays. We shall in all probability find this difference in the linkage
of the enzyme to the positive N-group. And if we assume the
enzymes to be conjugated of a co-enzyme and a apo-enzyme, the
difference between the enzymes is the consequence of different
apo-enzymes to which components the N-groups are linked.
Probabty, there are more than 2 different apo-enz 3 >-mes. This is
clear from the investigations published here and others being
published in a subsequent paper.

Summary.

1. A survey of the investigations, made earlier on the choline
esterase content in invertebrates, is a preface to this paper.

CUOLINE ESTERASES IN SOME MARINE INVERTEBRATES. ]49

2. The acetylcholine hydrolysing effect for different marine
invertebrates has been investigated.

3. The acetylcholine hydrolysing effect is compared with the
hydrolysis of two other choline esters, acetyl-/9-methylcholine and
benzoylcholine, wliich other authors have suggested using in
order to estimate the specificity of choline esterase. The results
show that the method is insufficient. Thus some species hardly
split either of the two substrates mentioned; this suggests the
possibility of the occurrence of an enzyme which is still more
specific as regards acetylcholine.

4. The properties of the choline esterases are discussed in the
light of present knowledge.

This investigation has been financially supported through the
award of a travelling scholarship of C. I*. Liljevalch jr.

I wish to express my grateful thanks to my wife for her valuable
help in collecting and preparing the animals.

References.

Alles, G. a., and R. C. Hawes, J. biol. Chem. 1940. 133. 375.

Alles, G. a., and R. C. Hawes, Science 1944, 100. 1b.

Ammon, R., Ergebn. Enzymforsch. 1935. 4. 102.

Ammon, R., Ibidem 1943. 9. 35.

Artemow, N. M., and R. N. Lurje, Bull. Acad. Sci. XJRSS. Ser. biol.
1941. 194. 278.

Artemow, N. M., and R. N. Lurje, Chem. Zbl. 1942. I. 1266.
Artemow, N. M., and R. N. Lurje, Chem. Abstr. 1943. 37. 2824.
Artemow, N. M., and R. L. Mitropoutansk,vja, Bull. Biol. Med.
exp. URSS. 1938. 5. 378.

Augustinsson, K.-B., Ark. Kem. Mineral. Geol. 1944. 18A. no 24.
Augustinsson, K.-B., Nature 1945. 156. 303.

Bacq, Z. M., Arch. int. Physiol. 1935 a. 42. 47.

Bacq, Z. M., C. R. Soc. Biol., Paris 1935 b. 120. 247.

Bacq, Z. M., Arch. int. Physiol. 1937 a. 44. 174.

Bacq, Z. M., Proc. Roy. Soc. Lond. Ser. B. 1937 b. 123. 418.

Bacq, Z. M., and D. Nachmansohn, J. Physiol. 1937. 89. 368.

Bacq, Z. M., and A. Oury, Bull. Cl. Sci. Acad. Roy. Belgique (5)
1937. 23. 891.

Bergmann, M., Adv. Enzymol. 1942. 2. 49.

Bissegger, a., and E. A. Zeller, Helv. physiol, pharmacol. Acta
1943. 1. C 86.

Bullock, T. H., and D. Nachmansohn, J. cell. comp. Physiol. 1942.
20. 239.

Corteggiani, E., and A. Serfaty, C. R. Soc. Biol., Paris 1939. 131.
1124.

150

KLAS-BERTIL AUGUSTINSSON.

Gautrelet, J., and N. Halpern, Ibidem 1935. IIS. 412.

Glasson, B., Pharm. Acta Helv. 1944. 19. 279.

Halpern, N., and E. Corteggiani, C. B. Soc. Biol., Paris 1935. 119.
1049.

Hawes, E. C., and G. A. Alees, J. Lab. clin. Med. 1941. 26. 845.
JuLLiEN, A., C. E. Acad. Sci., Paris 1939. 209. 1015.

Jullien, a., C. E. Soc. Biol., Paris 1941. 135. 1230.

JcLLiEN, A., and A. Bonnet, C. E. Acad. Sci., Paris 1941. 212. 813.

Jullien, A., and D. Vincent, C. E. Soc. Biol., Paris 1938. 129. 845.

Jullien, A., and D. Vincent, Ibidem 1941. 135. 956.

Jullien, A., D. Vincent, JI. Bouchet, and Vuillet, Ann. Physiol.
1938. U. 567.

Koschtojantz, Oh. S., Bull. Biol. Med. exp. UESS. 1936 a. 2. 34.
Koschtojantz, Ch. S., Ukrain. biochem. J. 1936 b. 9. 665.
Langemann, H., Helv. physiol, pharmacol. Acta 1944. 2. C 17.
Laubenfels, M. W. de. Science 1943. 98. 450.

Marnay, a., and D. Nachman.sohn, C. E. Soc. Biol., Paris 1937.
125. 1005.

Means, jr. 0. W., J. cell. comp. Physiol. 1942. 20, 319.

Mendel, B., Canad. Chem. 1943. 27. 608.

Mendel, B., Brit. chem. Abstr. 1944. A HI. 177.

Mendel, B., D. B. Mundell, and H. Eudney, Biochem. J. 1943.
37. 473.

Mendel, B., and H. Eudney, Ibidem 1943. 37. 59.

Mendel, B., and H. Eudney, Science 1943. 98. 201.

Mendel, B., and H. Eudney, Ibidem 1944. 100. 499.

Mendel, B., H. Eudney, and F. Strelitz, Nature 1944. loi. 737.
Mitropolitanskaja, E. L., C. E. Acad. Sci. UESS. 1941. 31. 717.
Nachmansohn, D., C. E. Soc. Biol., Paris 1938. 127. 894.
Nachmansohn, D., Bull. Soc. Chim. biol., Paris 1939. 21. 761.
NacThmansohn, D., and hi. A. Eothenberg, J. biol. Chem. 1945.
158. 653.

Narikashvili, S. P., Bull. Biol. Med. exp. UESS 1939. 7. 283.
Pennoit-De Cooman, E., Natuurwet. Tijdschr. 1940. 22. 62.

Eezek, a., and G. Haas, Biochem. Z. 1942. 312. 114.

Eichter, D., and P. G. Croft, Biochem. J. 1942. 36. 746.
Eiechert, W., and C. Schnarrenberger, Arch. exp. Path. Pharmak.
1942. 200. 225.

ScHAB-WuTRiCH, B., Helv. chim. Acta 1943. 26. 1836.

Tahmisian, Th. N., Anat. Eec. 1941. 81. Suppl. 122.

Vahlquist, B., Skand. Arch. Physiol. 1935. 72. 133.

A'^incent, D., and A. Jullien, C. E. Soc. Biol., Paris 1938 a. 127. 628.
Vincent, D., and A. Jullien, Ibidem 1938 b. 127. 631.

Vincent, D., and A. Jullien, J. Physiol. Path. gen. 1939. 37. 35.
Zeller, E. A., and A. Bissegger, Helv. chim. Acta 1943. 26. 1619.

Prom Mediciualco’s biological-chemical Laboratories, Copenbagen.

IiiYcstigatioiis ot* the Phosphatase Activity
ill Seruiii and Organs after Ligation of the
Coninion Bile Duct in Dogs.

INGER GAD.

Received 22 November 194.').

It lias long been known that obstructive jaundice both in man
and, produced experimentally, in animals is accompanied by an
increase in the serum phosphatase activity. In the course of time
various hypotheses have been forwarded in order to explain this
increase, but only three of them have gained common interest.

The first hypothesis was suggested by Roberts (1930, 1933),
who presumed that the serum phosphatase activity is increased
owing to the lacking secretion of phosphatase with the bile. Tliis
causes an accumulation in the blood. This hypothesis was forwarded
on the basis of investigations on the phosphatase content of the
bile, and it was supported, among others, by R. G. Anderson
(1935), Herbert (1935), Flood, Gutman and Gutman (1937)
and Gutman, Hogg and Olson (1940).

In the second place, there is Thannhauser’s et al. (1938),
assumption that the increase in serum phosphatase during jaundice
is not a true increase in the amount of enzyme, but rather an
activation of the phosphatase present, caused either by the for-
mation of activating substances or by the lack of inhibitors. As
the only support for this theory Thannhauser showed that icteric
serum after mixture with normal serum causes an activation of the
normal serum phosphatase, a result which Cantarow (1940), was
able to confirm, while other investigators failed to do so.

]52

INGEB GAD.

The third hypothesis was put forward by Bodansky (1937) who
stated that the increase in serum phosphatase activity as a con-
sequence of biliary obstruction is due to an increased delivery of
phosphatase from the liver tissue. BoDA>'SKy set forth the view
that variations in the serum phosphatase activity under patho-
logical conditions must be ascribed to disturbances in the phos-
phatase-production or -secretion in the organ which is suffering
in the respective case.

Among the hypotheses described above Bod.A'Nsky’s explana-
tion of the increase in serum phosphatase appears to be the most
probable one. Since, however, the phosphatase content of the liver
under normal conditions is not very high, a study of the phos-
phatase activity of the liver under icteric conditions ■vs'ill be neces-
sary in order to decide whether this hypothesis can be regarded as
well founded.

The following sections deal with the changes in the phosphatase
content of the liver, of other organs, and the serum, which arise
from a ligation of the common bile duct.

Analytic.al Method.

The phosphatase determinations were performed according to a
method developed partly on the basis of Bodansky’s (1933, 1937)
procedure, partly after Lundsteen and Vermeheen (1936).

0. 1 . ml serum or organ extract was applied; as substrate was used
disodium-^-glycerophosphate dissolved in an ammonia buffer, pH
9.95, containing SIg ions; 1 hour hydrolysis at 37°. The inorganic
phosphate was determined colorimetrically in the Pulfrich photometer
according to Brigg’s (1922) method.

Reagents.

1. 0.9% sodium chloride.

2. Substrate; 1 gm disodium-^-glycerophosphate, 7 ml n ammonium
chloride, 13 ml n ammonia, 2 ml n magnesium chloride, filled
up to 100 ml with distilled water.

3. 10% trichloroacetic acid.

4. An acid molybdenum reagent; 25 gm ammonium molybdate,
dissolved in 300 ml dist. water -p 75 ml cone, suplhuric acid,
diluted to 200 ml with distilled water. Imperishable.

5. 20% sodium sulfite. Durable c. 1 month.

6. 0.5% hydrochinone, 2 drops of cone, sulphuric acid per 100 ml.
Durable c. 14 days.

7. Standard phosphate solution: 0.4394 gm potassium phosphate
(Sorensen) per 1,000 ml. Contains O.i rag P per ml.

PIIOSPUATASE ACTIVITV IN SERUM ANP ORGANS.

153

1 ml serum (venous blood centrifuged for 20 min. nt 3,000 rev.) wns
diluted to 100 ml with 0.9% NaCl. Into a 10 ml centrifuge tube 2 ml
ft (2) were pipetted; 1 ml of the serum dilution was added,

e tube was closed with a stopper and was placed in the water thermo-
e ^ hour; subsequentlj”, the protein was precipitated by
1 'on of 2 ml trichloroacetic acid (3). A control was made by adding
immediately 2 ml trichloroacetic acid to serum dilution + substrate.
r'pn+ j loom temperature, the precipitated protein was

m supernatant 4 ml were pipetted off, 0.5 ml

Rnbif- ^ NaoSOs solution (5), 1 ml hydrochinone

.1 distilled water ad 10 ml were added. After 30 minutes,

e CO our intensity was measured in the Pulfrich photometer, filter

olank containing the same quantities of reagents as the
analysis m 10 ml dist. water (Fig. 1).

154

INGER GAD.

An extinction ciKve was measured with known amounts of phosphate.
No deviation from Lambert-Beer’s law was found within the measuring
range of the photometer.

The difference between the amount of phosphate in the analysis and
in the control represents the amount of phosphate liberated by the
phosphatase. The phosphatase unit is defined as the quantity of phos-
phatase which liberates 1 mg of phosphorus from the substrate in the
course of 1 hour at 37°.

The determination of the phosphatase activity in organic extracts
was performed on exactl)^ the same lines as described for serum.

The extracts were prepared in the following way. 5 gm of fresh organ,
if possible an average sample (freed from tendons, membranes, blood
etc.), were ground with sand and extracted with 50 ml dist. water
for c. 10 minutes and, subsequently, the extract was centrifuged. 1 ml
of the extract was diluted to 10 ml with 0.9% NaCl. For the phosphat-
ase determination 1 ml of this dilution was used in the same way as
described for serum. Extracts which are especial!}' rich in phosphatase,
such as intestine- and kidney extract, were further diluted 10 times
prior to the phosphatase determination, so that in no case more than
10% of the amount of glycerophosphate used was split, since the cleav-
age products formed during hydrolysis have some inhibiting effect if
more than the mentioned 10 % of the total amount of the substrate are
hydrolyzed (Bodansky 1933, 1937).

This method for the production of organ extracts was completely
satisfactor}' for the comparison of the phosphatase content in normal
and in icteric organs.

The phosphatase activity was calculated from these comparisons for
extracts which are 1% with regard to dry matter. The content of dry
matter was determined on the first aqueous extracts before dilution
with NaOl by drying to constant weight at 100°; the dry matter was
always found to be between 1 and 2%.

The results of the phosphatase determinations obtained by this
method can approximately be compared with those of Bodansky
(1933, 1937) since the definition of the phosphatase unit is the same.
However, the present values appear to be somewhat higher, as I have
used a more alkaline substrate and addition of Mg ions.

Animal Material.

All the experiments were carried out on healthy and adult dogs.

For some time previous to and during the whole experimental period
the animals were kept on a constant diet of black rye-bread and milk.
No significant loss in weight could be observed in the course of the
experiments.

Experiments.

1. Determination of Normal Values.

A series of determinations was performed of the phosphatase
activity in serum and organs of normal dogs.

I'lIOSI’llATASK ACTIVITY IN SERWM AND ORGANS. 155

Table 1.

||M

Pancri'.-is

Kitlnoy

cortfx

Njilecn

Iiitc-stino

(jcjiiiium)

3.1

ll.r

l.u

3.0

0. 1

4.2

o.s

fj.ri

27..S

100

■).'50

3.y

7.:t

.T3.n

12!)

•19S

ri.o

5.7

fiG.o

r>«2

3.n

17.-.'

3(5.2

J20

■lOfS

2.0

5.;

H.’i.u

(5(5

2.3(5

G.'.i

(5.7

.3S.T

19.8

{•u3()

G.l

22.7

147

1C,.-;

478

2,7

4.r.

23.0

f.r»

42S

8.1

13.0

215.11

17(5

7.4

(58(1

J.s

H.:,

.30.7

ir,o

7.0

487

3.1

30.H

lift

.31 f)

2..S

3..-,

.3a.:t

13)

9.3

175

jut 100 ;nl >>crom ninl jicr 100 ni! orijmi cxtr.icl willi 1
<lry lu.im-r. pH 9.'J3. 1 hour hydrolj^ia nt 37'.

T!ic table revoal.s tliat ihc info-stino contnin.'; by far the greatest
utiiount of phosplmlase; tiien follow the leiJney cortex, jianeroa.'s.
and finally liver and .‘spleen, both .•showing rather low plio.sp])nta.se
contents. Thc.se rcstilt.*! are in good ngreeinenf- with those previoiislv
found (I'oLhKY and Kav 10:JG).

The table Hhow.s, moreover, that the phosphatase activity’ in the
.same organ.s varies to .some degree from one animal to another.
However, the variation was not considered .sufficiently large to
prevent the application of this normal material being used as a
basis for the comparison with the pho.spliatasc valuc.s obtained
after ligation of the bile duct.

2. Phosphataao Activity after Ligation of tho Common

Bilo Duct.

Ligation of tho common bile duct was carried out on 9 dogs.
Before tlie operation, wliicli was performed in morpliine-etber
nnrco.si.?, the Berum piio.sphntasc activity was cheeked for two
day.s. In .some cnsc.s, a jiicce of tlie liver was removed (an electric-
ally lieated thcrino-cnutcry was u.scd, no blecding.s occurred) for
a phosphatase determination simultaneously with the ligation.
The ictcru.s indc.v was determined by means of MEunKN 0 UACHT',s
method.

T to 19 day.s after the operation the clogs were killed.

156

INGER GAD.

Table 2 shows the results of the determination of serum phos-
phatase activity and icterus index after ligation of the common
bile duct.

Talile 2.

Serum phospliaiase activity and icterus index after ligation
of the common bile duct (dog).

Dog Ho.

Before

operation

2 days after
operation

4 days after
operation

7 days after
operation

ict. ind.

phosph.

9.7 kg

5.8

58.0

86.0

27.0 »

2.5

5.4

35.5

109.0

23.5 »

3.5

3.4

48.0

77.5

19.0 »

3.0

1.9

22.0

78.0

30.0 »

2.0

3.9

12.0

54.0

52.0

21.0 »

3.5

3.9

7.5

8.4

60.5

123.0

10.0 »

3.0

3.5

44.0

106.0

141.0

19.5 »

3.2

5.2

27.0

88.0

128.0

21.5 1)

2.5

2.9

67.0

119.0

Phosphatase units per 100 ml serum. pH 9.05. 1 hour hydrolysis at 37°.

The table shows that, in the course of some days, the serum
phosphatase activity increases simultaneously with formation of
jaundice. The rate at which these changes occur is quite individual
since, for example, 2 days after the operation the icterus index
observed varies between 7.5 and 48, and the serum phosphatase
activity ranges from 8. 4 to 67 units. 4 days after the ligation the
figures were found to be between 10 and 58 for the icterus index
and between 54 and 119 units for the phosphatase activity,
respectively; It is, furthermore, clear that the increase in the icterus
index and in the serum phosphatase activity do not run parallel.
The results are in good agreement with those obtained by other
investigators.

With respect to dog No. 5 the experimental period was extended
to 19 days. It was found that the phosphatase activity in serum
reached a maximum of 116 units in the course of 10 days and then
began to decrease again; after c. 3 weeks it had fallen to 46 units.
Also these observations are in good agreement with earlier results.

Table 3 shows the results of phosphatase determinations on
organ extracts after jaundice of various durations.

The table shows that bile obstruction has quite a different effect
on the phosphatase activity in the organs. In the liver, for example,
an increase to c. 10 times the normal value occurs (calculated on

PHOSPHATASE ACTIVITY IN SBHHM AND ORGANS.

157

Table 8.

Phosphatase activity in serum and organs before and after ligation
of the common bile duct {dog).

Dog No.

Duration

Serum

Liver

Pancreas

Kidney
cortex .

Spleen

Inte.stine

(jejunum)

icfcenis

be-

fore

after

be-

fore

after

4. 19 kg

4 days

1.9

3.9

68.5

9. 21 »

4 »

2.5

119

48.0

33.5

224

19.4

1830

3. 23 «

5 »

3.4

9.7

60.2

6. 21 »

3.9

123

102.5

51.4

200

8.0

650

7. 10 »

7 »

d.5

141

52.0

30.5

234

10.7

1350

8. 19 »

7 *

5.2

128

32.0

131

7.3

1465

5. 30 »

19 I

3.9

4.2

12.8

26.8

395

Normal values

1.9-

-8.1

3-5

—17.2

22.7—56

65—198

7.1—17

175—686

(Table 1)

Phosphatase units per 100 ml serum and per 100 ml organ extract with 1%
dry matter. pH 9.93. 1 hour hydrolysis at 37°.

the basis of the average figures) during a short-lasting jaundice.
In the intestine an increase of c. 300% was observed; measured
in absolute figures this elevation is very remarkable.

In the case of pancreas and spleen no increase in the phosphatase
activity content was found, while the phosphatase activity of the
kidney may appear insignificantly increased.

The results obtained after long-lasting icterus (dog No. 5) indi-
cate that not only the serum phosphatase as mentioned previously,
but also the liver and intestinal phosphatase activities decreased
after a maximum to almost normal values.

Discussion.

It results from the experiments that the ligation of the common
bile duct in dogs not only causes an increase in the serum phos-
phatase activity, a phenomenon which has been known earlier,
but simultaneously produces an increase in the phosphatase con-
tent of liver- and intestinal tissues.

Some interdependence seems to exist between the phosphatase
activity in serum and in these organs during obstructive jaundice.

The increased serum phosphatase may thus be ascribed to an
overproduction of phosphatase in the liver, a hypothesis primarily
assumed by Bodansky 1937. It is, however, just as probable —
or rather more probable — that the increased serum phosphatase
during obstructive jaundice originates from the intestine which.

158

INGER GAD.

both normally and under icteric conditions, may give off its phos-
phatase to the liver, functioning as a kind of regulator for the
phosphatase of the plasma.

More detailed investigations of the significance of both liver and
intestine for the increase in the serum phosphatase activity after
ligation of the common bile duct will be described in papers under
preparation.

Summary.

1. On the basis of Bodakskv’s and Lukdsteen and Ver-
mehren’s procedures an analytical method is described for the
determination of the phosphatase activity in serum and organ
extracts.

2. The phosphatase activity is determined in a series of organs
of normal dogs. The intestine was found to have the highest phos-
phatase activity, then follow kidney cortex, pancreas and, finally,
liver and spleen which both have rather low phosphatase contents.

3. The observations of other authors concerning the serum
phosphatase activity after ligation of the common bile duct have
been confirmed insofar as, a few days after ligation, a jaundice
is observed which is accompanied by an increase in the serum
phosphatase activity of up to c. 40 times the normal value. No
parallelism is seen between the intensity of the jaundice and the
serum phosphatase activity, furthermore an increase in the phos-
phatase activity of the liver- and intestinal tissues is observed,
while the stasis of the bile does scarcely affect the phosphatase
contents of pancreas, spleen and kidney.

4. It is assumed that the increased serum phosphatase during
obstructive jaundice may originate from the intestine.

I am indebted to cand. pharm. V, Larsen for performing the
operations of the dogs.

Literature.

Anderson, B. 6., St. Barth. Hosp. Bep. 1935. 68. 221.

Bodansky, a., J. biol. Chem. 1933. 101. 93.

Bodansky, a.. Ibidem 1937. 120. 167.

Bodansky, A., Enzymologia 1937. 3. 258.

Briggs, 0., J. biol. Chem. 1922. 53. 13.

Cantarow, a., Amer. J. Clin. Path. 1940. 10. 858.

Flood, C. A., E. B. Gutman, and A. B. Gutman, Arch, intern. Med.
1937. 59. 981.

PlIOSPUATASE ACTIVlir IN SERUM AND ORGANS. 159

Flood, C. A., E. B. Gutman, and A. B. Gutman, Amer. J. Pliysiol.
1937. 120. G9G.

Folley, S. J., and H. D. Kay, Ergcbn. Enzymforsch. 1930. 5. 159.
Gut.man, a. B., B. M. Hogg, and K. B. Olson, Proc. Soc. Exp. Biol.,
N.Y. 1910. U. G13.

Herbert, F. K., Brit. J. Exp. Path. 1935. 16. 3G5.

Lundsteen, E., and E. Vermehren, G. E. trav. lab. Carlsberg 1936.
21. 147.

Roberts, W. M., Brit. J. Exjl Path. 1930. 11. 90.

Roberts, W. M., Brit. Med. J. 1933. 3773. 734.

Thannhauser, S. j., M. Reichel, J. F. Grattan and J. F. Maddock,
J. biol. Chem. 1938. 121. G31.

From the Biochemical Institute, Aarhus University, Denmark.

Studies on Serum Pliospimtase Activity in
Relation to Experimental Biliary
Obstruction in Rabbits. 11.

J0RGEN HOFFMEYER, OLAF JAELING and FRITZ SCH0NHEYDER.
Received 25 Novemlicr ID'IS.

In a previous publication (Jallixg, Laursex and Yolqvartz,
1946) it tvas shown that the increase in the serum phosphatase
activity in relation to experimental biliary obstruction may partly
be explained by retention of phosphatase which is normally ex-
creted with the bile, but besides a delivery of phosphatase from
one or more organs may be of importance.

It is a natural thought, that the content of phosphatase is
changed in the organ or the organs which deliver phosphatase to
the blood. The experiments on rabbits which are presented in this
paper show, that it is not possible to demonstrate any certain
change in the phosphatase content of liver tissue or the intestinal
mucosa at the moment when the increase in serum phosphatase
after ligature of the common bile duct is Imown to be maximal.
In the case of the renal cortex, on the other hand, it is possible to
demonstrate a significant decrease in the phosphatase activity in
connection with obstructive jaundice, as has been previously
established by Takata (1932). This fall in the phosphatase activity
of the renal cortex is without any causal relationship to the in-
crease in serum phosphatase, the hyperphosphatasemia in animals
with ligature of the bile duct and removal of both kidneys being
accompanied by an increase in serum phosphatase of the same size
as in animals where the kidneys had not been removed. Our deter-
minations of the content of alkaline phosphatase in liver tissue

SEKHM PHOSPHATASE ACTIVITY.

161

and intestinal mucosa have thus not brought us any nearer to the
solution of the problem concerned with the serum phosphatase
increase in connection with biliary obstruction in rabbits.

Method for the Determination of Tissue Phosphatase.

Kay (1926) and Macfarlane, Patterson and Eobison (1934) have
determined phosphatase in tissue using glyccrophosphatasc as substrate.
When their procedure is followed using disodium phenylphosphate as
substrate no proportionality between amount of tissue used and phos-
phatase activity measured could be demonstrated. The following pro-
cedure was therefore worked out.

Immediately after the animal has been killed the organs in which
the phosphatase content is to be determined are taken out. In case of
the kidney the renal cortex is isolated. In case of the intestine 50 cm of
this organ beginning 10 cm distally from pylorus are isolated and opened
with a pair of scissors. The mucous membrane is rinsed with tajr water
and wiped and is then scraped off.

Portions of 0.25 to 0. a g of the comminuted tissue arc weighed out
and ground with fine sand and a Icnown amount of glycerol (87 per cent),
which is added in small portions during grinding. This is continued
until the mixture docs not contain any visible tissue particles. The
determination of phosphatase is carried out not later than two hours
after the beginning of the grinding, and the glycerol extract is carefully
stirred before adding samples of it to the buffer substrate solution.
Phosphatase is estimated according to Buon and Btjcn’.s method (1939)
for the determination of serum phosphatase, the Pulfrich ijhotometcr
and a standard curve given by Jalung, Laursen and Volqvartz
(19'45) being used. Buch and Buck’s unit refers to 50 ml of serum, 0.26
ml of scrum being used for the determination. The flwsjthatasc activity
of tissue is calculated as •phosphatase units per g of dry tissue dispers<d
in 50 ml of glycerol, using 0.25 ml glycerol extract for the determination.
The following example illustrates the procedure and the calculation of
the phosphatase content in tissue.

liabhit No. 37. Immediately after the death of the animal the liver
is taken out. A part of the organ is comminuted. 8 samples of about
0.5 g are weighed out, and each sample is ground with sand and 20 g
of glycerol. On each glycerol e.xtract duplicate determinations arc per-
formed with O.io, 0.15, 0.20 and 0.25 ml of extract. In order to get the
same volume in each buffer-substrato-enzymo mixture 0.15, O.io and
0.05 ml of glycerol are added to the test tubes, to which are subsequently
added 0. i o, 0. 1 6 and 0. 2 o ml of extract. In Table I (Columns 3 to G) are
given phosphatase units per 50 ml of diluted glycerol extracts, computed
as the mean value of duplicate determinations. When the phosphatase
activities found for O.io, 0<15, O .20 and O .25 ml extract are plotted in
relation to amount of extract used a straight line is obtained for each
tissue sample; hence it is possible to compute the phosphatase activity

11 — ' 1 GO 215 . Acta phys. Scavdinav. Vol, 11.

162 J0RGEN HOFFMEYER, OLAF JALLFNG AND FRITZ SCH0NHEYDER.

TftWo 1.

Determination of ihe Phosphatase Content in 8 Samples of the
Same Liver Tissne.

Sample

Liver

tissue

Phosphatase units per 50 ml of
diluted glycerol extract

Ph. units per
g dry tissue in

*0.10

♦0.15

*0.20

*0.25

50 ml glycerol

0.4990

20.8

36.0

43.9

57.5

144.2

0.4997

21.0

37.5

45.3

59.9

149.4

0.4995 ’

19.3

34.0

42.0

49.7

133.0

0.4995

21.9

33.3

39.9

54.4

137.8

0.4998

19.3

33.9

47.3

53.0

155.2

O.S003

22.9

33.4

43.2

53.5

141.2

0.4994

17.3

32.9

40.7

49.7

127.C

0.4998

22.8

36.1

40.7

55.3

143.4

Percentage of dry tissue: 26.20

/o*

Average 141.48

* ml of undiluted glycerol extract in 0.25 ml dilution.

of tte tissue as tlie average of all determinations after liamng corrected
for the dilution. The observed proportionality is also found for extracts

of intestinal mucosa and kidney tissue ■when the following proportions
are used: 0.3 g intestinal mucosa to 15 g of glycerol and 0.25 g kidney
tissue to 30 g of glycerol. The phosphatase units per g of dry tissue
dispersed in 50 ml glycerol = A is computed by insertion into the for-
mula: A = 2 X where n = phosphatase activity per 0.25 ml

Q. X V X t

undiluted extract, a = amount of glycerol used in g, d = density of
glycerol, v — weight of tissue in g, t = dry substance in per cent,
calculated by drying two samples of tissue to constant weight at 105° C.

It appears from Table I, that there is a fairly good agreement between
the numbers of phosphatase units per g dry matter in 50 ml of glycerol
calculated from the 8 samples of tissue.

In the experiments which are presented in what follows the phos-
phatase determinations are generally performed on 2 extracts ficom the
same organ. On each glycerol extract duplicate determinations are per-
formed on 3 different dilutions. The inaccuracy of the method has been
determined on the basis of all the determinations which were carried

out in duplicate or more. When the determinations for each organ were
pooled the following coefficients of variability were found, liver tissue:
V = 3.97 %, intestinal mucosa: V = 5.28 %, Mdney tissue: V = 3.92 %.

Experimental.

Our animal material consisted of normal rabbits and rabbits
with ligature of the common bile duct. All the animals weighed
between 2 and 3 kg. The controls had normal values of serum phos-
phatase, viz. 3 to 12 units, as was also the case with the operated

SERUM PflOSPITATASE ACTtVITT.

Table II.

Phosphatase Activiii/ in Kidney Tissue.

163

Normal rabbits

Rabbits Tvitb ligature of tho
common bilo duct

Babbit

No.

Kidney tissue

Babbit

No.

Serum

pb.

units

Kidney tissue

Dry

matter

• Ph.

units

Average

Ph.

units

Dry

matter

* Pb.
units

Average

Ph.

units

22.7G

604.4

460.7

482.0

41 '

134.2

24.37

231.0

247.0

239.0

23.18

468.7

453.0

461.2

128.4

23.02

293.4
292.0 i

292.7

35 j

23.33

319.5

338.0

329.1 j

143.5

25.59

277.3
298.5 j

287.9

25.23 j

304.2

304.2 1

349.1 j

25.38

310.4

314.0

312.2

135.0

22.05

334.4

320.9

334.8

364.2

320,4

22.80

359.0

3G2.9

132.0 i

21.82

335.5

328.0

365.4

329.8

23.40

467.8

502.3

488.7

99.5

24.53

155.0

151.4

163.7

23.07

23.80

24.35

21.51

495.0

447.2

452.2

418.2
399.5
466.4

443.3

311.0
307.9

449.7

408.9

454.9

309.9

68 j

142.5

115.5

116.3

93.0

23.40

23.00

21.24

24.18

316.2

338.9

291.5

272.7

258.1
266.0

! 327.1

328.2

327.0

282.1

262.1

327.7

23.05

387.1

* Ph = Phospimtaso units per g of dry tissue dispersed in 60 ml glycerol
Normal rabbits: Mean value of pb.: 39G.0; a = 70.0.
Operated rabbits: Mean value of ph.: 283.7; a s= 55.7.

animals before tbc operation. The animals with ligature of the
common bile duct were killed 17 to 24; hours after operation. As
shown by Jalling, Laursen and Volqvartz (1946) the increase
in serum phosphatase after ligature of ductus choledochus is
maximal 17 to 20 hours after the operation. Immediately before
the death of the animal the serum phosphatase activity was de-
termined, and the values found are given in Tables II — IV. In all
the animals ivith ligature of the bile duct a marked jaundice was
demonstrated. At section tho gall bladder was found to be very
distended. A few animals in which only moderate or no increase in
serum phosphatase after the operation was found have been ex-
cluded from the material.

164 J0RGEN HOFFMETEB, OLAF JALLING AND FRITZ SCH0NHEYDER.

Table HI.

Phosphatase Activity in Liver Tissue.

Normal rabbits

Rabbits vlth ligature of the
common bile duct

Liver tissue

Eabbit

No.

Dry

matter

*Ph.

units

Average

Ph.

units

Rabbit

No.

ph.

units

Dry

matter

»Ph.

units

Average

Ph.

units

26.20

27.62

25.28

26.34

28.69

28.72

28.53

26.03

28.34

(See

Tabl.1)

72.6
72.3

172.7

159.6

41.7
40.2

40.2

63.8

56.1

52.8

68.3
66.0

74.3

76.8

66.2

69.4
81.1
86.2
82.6
85.1

141.5

72.5

166.1

40.7

54.2

67.2

75.6

67.8

83.7

83.9

163.5

143.8

134.2

135.6

132.0

99.5

142.5

115.5

116.3

93.0

27.70

21.28

23.38

24.34

24.27

25.14

24.45

23.80

24.03

23.62

211.8

71.5

66.9

78.3
75.7

88.4

88.5

91.1

103.1
108.4

104.1

45.2

47.5
140.9
131.8

67.9

67.4
68.0

66.5
55.1

59.9

211.8

69.2

77.0

89.3

105.2'

46.4

136.4

67.7

67.3

57.5

* Ph. = Phosphatase units per g of dry tissue dispersed in 50 ml of glycerol.
Normal rabbits: Mean value of ph.; 85.3; a = 38.8.

Operated rabbits: Mean value of ph.: 92.6; u = 49.3.

In tlie animals determinations of the phosphatase content have
been performed on kidney tissue, liver tissue and intestinal mucosa
according to the method described above and the results are re-
corded in Tables II, III and IV.

It is remarkable that the phosphatase activities in the examined
tissue from liver and intestinal mucosa vary extremely from one
animal to the other. The standard deviation is somewhat smaller
in the case of kidney tissue. At the time where the hyperphosphatas-
emia is maximal there is no doubt that the phosphatase activity
in the renal cortex of the animals with obstructive jaundice is
smaller than in the control material. Assuming the validity of the
zero-hypothesis and calculating the common a = 64.6, the value
t =4.03 is obtained. iProm Fisher’s Table of t it is seen that

165

operated xa '

^oriu‘“ . , v,ls;X>i.ea“ '• —

operated xa ^l.at tltC diffctcnCC tS

0 001 It » owo couteiit Ot *0

-pltos^atasc ^,atasc.

p ^ 0.001. 1. - ^Wiatasc <^«-;-;^^,,tasc.

“ L inoi«“ <>'>5"„ i„Mrto»o» to *0 l,„„n es-

*'”“’"t!«raltei«6<'t«t»-''*"

p^oBp^tatase

166 J0RGEN HOFFMETER, OEAF JALEING AND FRITZ SCn0KHEyDER.

tablislied by tbe following experiments. On Eabbits 37 and 38
doublesided nephrectomy was carried out in connection with liga-
ture of ductus choledochus. The values of serum phosphatase are
to be found in Table V.

Table Y.

Serum Phosphatase in 2 Pabbits after Ligature
of Ductus Choledochus and Doublesided
Nephrcctomp.

Babbit No. 37

Rabbit No. 38

Time after

Scrum

Time after

Scrum

operation

phosphatase

operation

phosphatase

hours

units

hours

units

2V.

33.0

3V.

82.8

73.5

7V.

102.5

liv.

98.2

12V.

131.1

14'A

143.5

18V.

159.0

22V,

162.0

It is seen that the rise which can be observed in the rabbits is of
the same order of dimension as in rabbits in which the kidney has
not been removed (cf. Jaeeing, Laiirsen and Voeqvartz, 1945).

Takata (1932) has stated that the /5-glycerophosphatase con-
tent in kidney and liver from rabbits with experimental obstructive
jaundice was lower than in normal rabbits. The animals were killed
48 hours after ligature of ductus choledochus (serum phosphatase
determinations were not carried out). WTiereas the difference
between the phosphatase contents in kidneys from operated
animals and controls is significant, Takata’s material does not
allow any conclusion with regard to the liver. Takata maintains
that the lowered phosphatase contents in organs in connection
with obstructive jaundice may be partly ascribed to the cholic
acids which have entered the blood circulation, injections of
sodium cholate subcutaneously being able to diminish the phos-
phatase content, which beforehand had been decreased by obstruc-
tive jaundice. In our opinion, however, the question concerned
with the low phosphatase activity in the kidneys of rabbits with
obstructive jaundice needs further examination, as we have no
information about which kind of cholic acids are accumulated and
their concentration, in the kidneys from operated animals.

SERUM PHOSPHATASE ACTIVITY.

167

The determinations of phosphatase in liver tissue and intestinal
mucosa show that the values for these tissues from rabbits ivith
obstructive jaundice lie within the range of normal values. There
is no doubt that liver and intestinal mucosa, perhaps other organs
too, contain so much phosphatase that they are able to deliver
phosphatase to the blood, without any change being demonstrated
in the organs by means of the experimental technique employed.
The large spreading in the organ phosphatase content complicates
the solution of the question further. It must at present be con-
sidered out of the question that a real increase in the phosphatase
content of liver and intestinal mucosa should occur in immediate
connection with ligature of the bile duct in rabbits.

Snmmary.

The studies concerning the mechanism of the development of
hyperphosphatasemia in rabbits vdth obstructive jaundice have
been continued.

1. A method is described for the quantitative determination of
phosphatase in tissues.

2. By means of this method determinations of phosphatase
have been carried out on liver tissue, intestinal mucosa and renal
cortical tissue from normal rabbits and rabbits on which ductus
choledochus has been ligated. 17 to 20 hours after the operation,
when the hyperphosphatasemia is maximal it is not possible to
demonstrate any difference in the phosphatase contents of liver
tissue and intestinal mucosa. In renal cortex a significant fall in
phosphatase has been demonstrated.

3. The investigations do not give any explanation of the hyper-
phosphatasemia in rabbits in connection with obstructive jaundice.

This work has been aided by a grant from the »P. Carl Peter-
sen’s Fond)).

References.

Buck, I., and H. Buen, Acta med. Scand. 1939. 101. 211.

Jalling, 0., T. Laursen and K. Volqvartz, Acta physiol. Scand.
1915. 10. 70.

Kay, H. D., Biochem. J. 1926. 20. 791.

Macfarlane, M. G., L. M. B. Patterson and R. Robison, Biochem. J.
1934. 28. 720.

Takata, H., j. Biochem. 1932. 16. 83.

From the Physiological Department, Karolinska Institutet, Stockholm.

The Presence of a Substance with SympathinE
Properties in Spleen Extracts.

D. S. T. EULER.

Received 27 November 1945.

Since the discovery by Loewi in 1921 of the liberation of an
adrenaline-like substance on stimulation of the accelerator nerves
of the heart evidence has accumulated to show that probably all
adrenergic nerves owe their effect to some special substance pro-
duced or liberated at the endings of these nerves. As to the ac-
tive principle liberated from the heart, or obtained in extracts
thereof, Loewi found that it conformed in its biological actions
and chemical properties with adrenaline. On the other hand, the
work of Cannon and his co-workers (1937) on the biological ac-
tions of “sympathin” from various sources appears to show quite
conclusively that the mediator substance on several occasions
is not adrenaline.

The problem merits special interest in view of the numerous
reports dealing with the occurrence of blood pressure raising
extracts of various organs, apart from the adrenals, the hypo-
physis and the kidney, where specific vaso-active substances are
responsible for the actions.

The first observations seem to originate from Oliver and
Schafer (1895) who found pressor actions in aqueous spleen
extracts after a primary fall in blood pressure. Their observations
were confirmed by Bingel and Strauss in 1909. Vincent and
Sheen (1903) have also observed pressor actions in organ extracts,
and later such actions have been reported by Roger (1922) and
James, Laughton and Macallum (1926) in liver extracts. In
an extensive study Collip (1928) was able to show that extracts

B a««*y^we of

c otgai^ con- soluble lU aud

coto “ ®tto aof«' '*'SSy «toWo to “irgtoui a !«"

’“rS r to ^e'a '■"

acetone coca^ue. ^sttatea

“"'oS^^ce ias ’>»“ V®. activity to^^^

a®® ‘Stty toatotyto”* S«v

TtAB'ENVica aud IS t

■g^^vjER an Q5,osST.tA3^ ( .^flalcb they piesaot sub
pteasot su^^^^^^ .lltteualn (a-g^^

anadieuabne ^^eactibed . ^ xeuiu ot ^TP ^ liypet

.^_,„v,-o (1942) b A^fffiTeut ftc w Aettaiu I’TP -nptties

,t5 on mantua^- 193d) T r^weaUbet-

Ueaaltoe jet (,ee adteactP^^

UW^ au4 K s COT ftog'a ^ fitees ol

T'f S toil f o\^“t Uacta ol ato^^f («S9) aai
i^cit laa osaamcd to

:SSai w» atoaaato- l^o'^^^cso-activc

cfJeeOT to tave i-- ofV ^

“I ooattoaatiou “* ‘^audtltoaav^'‘J;„jcrtaBCc to^
autotaac«ia^^;l^ion, » “0® essot

170

tr. S. V. EULER.

Fig. 1. Blood pressure, cat, chloralose.

1. Extract 0.125 g cattle spleen.

2. 0.25 g of same, treated with N WaOH at 100“ C for 10 minutes.

3. (After 10 mg/kg cocaine hydrochloride i. m.) 0.5 f‘g adrenaline.

4. 0.06 g of c.xtract 1.

6. Same, treated as in 2.

C. 0.2 g of same.

it was announced that extracts from a variety of organs — ex-
cept placenta — contain unexpectedly high amounts of pressor
activity of a land similar to that of adrenaline. The present paper
is concerned with some experiments made in greater detail with
extracts from spleen which w'as specially rich in the pressor sub-
stance.

Experimental.

1. Preparation and testing of extracts.

Extracts were made from fresh spleen of cattle. To the ground tissue
2 volumes of ethanol and 2 ml 10 N H-SOi per kg tissue was added and
the mixture left for 1 — i hours in room temperature under occasional
stirring. After filtering and evaporation of the filtrate to a small volume
fatty material was removed with ether. When tested on the cat’s blood
pressure under chloralose the aqueous extracts caused a fall in blood
pressure followed by a more or less conspicuous rise (Eig. 1). Pre-treat-
ment of the animal with ergotamine tartrate (Gynergen)^ in a dose of
0.1 mg per kg generally made the pressor response appear much more
distinct. This effect is due to the exclusion of the mechanical presso-
regulation mechanism (Euler and Schjiiterlow, 1945). Sometimes
vagotomy increased the pressor responses. Babbits were unsuitable

* Kindly put at my disposal by Messrs Sandoz A.-6., Basel.

SYMPATHIK E PROPERTIES IN SPLEEN EXTRACTS.

171

Fig. 2. Blood pressure, cat, chloraloso.

1. 0.2 lipid ether extract of fresh cattle spleen.

2. 0.2 lipid ether extract of spleen stored at room temp, for 5 hours.

3. 0.3- lipid ether extract of fresh liver,

4. (After 10 mg/kg cocaine hydrochloride i. m.) 0.1 of extract I.

6. 5 fig adrenaline.

6. 0.1 of extract 2.

7. 0.2 of extract 3.

8. 0.2 lipid ether extract of liver stored at room temp, for 5 hours.

for the test since they mostly reacted "with fall in blood pressure
only, unless highly purified extracts were used.

A question of primary importance was whether the time of prepara-
tion of the organ after the death of the animal had any influence on
the yield of pressor activity especially in view of the findings of Grabe,
Krayer and Seelkopp {1934) on pressor actions in liver extracts.
It was found that the most active preparations were obtained when
the organ was taken out immediately after the slaughtering and minced
in cold acid alcohol or stored for a few hours only in the refrigerator
before the extraction. Organs left in room temperature for some hours
yielded less pressor activity (J'ig. 2).

2. Puiiflcation of extracts.

A disturbing factor was the depressor action regularly preceding the
pressor response. Several attempts to separate the depressor principles
from the pressor were made by means of treatment witb various or-
ganic solvents such as alcohol, acetone and ether. It was noticed that
the pressor substance or substances were to some extent soluble in
mixtures of alcohol and ether, but, unfortunately, this also applied to
the depressor effect. The best results were obtained by shaHng tbe
aqueous extract after evaporation of the alcohol with a solution of
organ lipids in ether and subsequent shaking of the lipid-ether with

172

TJ. S. V. EULER.

Fig. 3. Blood pressvire, cat, chloralose.

1. Sublimate precipitate from lipid-ether extract of cattle spleen.

2. Sublimate filtrate.

3. 1 ftg adrenaline.

4. Same as (2), heated to boiling with N NaOH.

5. Half the dose of (2), untreated.

6. 1 /<g adrenaline heated to boiling -with N NaOH.

7. (After cocaine hydrochloride 10 mg/kg i. m.) 1 /ig adrenaline.

8. Same as (2).

9. (After ergotamine tartrate 2 mg/kg i. v.) 5 times the dose in (2).
10. 5 //g adrenaline.

5 — 10 % sodium sulphate solution. The sodium sulphate was removed
by addition of 3 volumes of alcohol and the filtrate, after evaporation
of the alcohol, showed a good pressor effect but only little depressor
action (Fig. 7). By this treatment a considerable purification was
attained, and a number of biological tests could be performed on the
purified extract. These extracts will be termed lipid-ether extracts.
After concentration and renewed extraction with lipid-ether there
was practically no depressor action left. A further purification was
obtained by treatment with HgClj. After complete precipitation with
HgCla in alcohol and sodium acetate the filtrate was decomposed with
HoS and, after removal of the HgS, concentrated, neutralized with
NaOH, evaporated to a small volume and taken up in methanol,
filtered and freed from alcohol. Treatment with HgClj removed all
of the depressor activity occasionally left in the lipid-ether extracts
with the precipitate, from which it could be recovered by decomposi-
tion with hydrogen sulphide, as will be seen from Fig. 3. The pressor
effect appears to be unchanged as to its general type though some loss
occurred and also some changes in stability (see under i).

3. Biological actions of extracts.
a. Blood 'pressure.

The lipid-ether extracts and the HgCl.-filtrates were tested
on the blood pressure of the cat and the rabbit. The responses on
the rabbit in urethane anaesthesia were, on the whole, rather

Fig. i. Blood pressure, cat, chloralose.

1. 2 /<g dihydroxy-nor-ophedrino (D. N. E.)

2. 2 /(g adrenaline.

3. 0.02 purified extract of cattlo spleen.

4. (After 2 mg orgoto.xino per kg) 10 /<g adrenaline.

.■i. 10 1% D. N. E.

0. 0.1 (3).

7.-9. Same as 4—0 after 3 mg crgotoxinc per kg.

poor, though a pure action was mostly obtained. On the cat, on
the other hand, pure and strong pressor responses were consis-
tently recorded. The pressor action of the extracts w'as of the same
general type as that of adrenaline and was accompanied by an
acceleration of the heart. Certain differences in action were noticed,
however, which were in some instances so marlced that it became
highly doubtful whether tlie active substance was identical wdth
adrenaline, in spite of other observations indicating a, near rela-
tionship to this substance. Thus, in many animals, adrenaline
caused a ‘step’ (Fig, '3) or a depressor notch whereas an action of
this type was never observed with the spleen extracts. Further,
the rise in blood pressure usually occurred quicker with the latter
as seen from the figure. Usually the increase in heart frequency
was more marked mth the extracts than ndth adrenaline.

Effect of cocaine. The enhancement of the adrenaline action
after a small dose of cocaine, as discovered by FRonucir and
Loewi (1910), also applied to the pressor action of the extracts.
(Fig. 1, 2, 3) This strongly indicated an adrenaline-like body,
since the cocaine effect appears only with certain dioxyphenol

174

U. S. V. EULEE.

derivatives. The increase in action was not always parallel in ex-
tracts and with adrenaline, but varied in different extracts, ac-
cording to the degree of purification and previous treatment.
Thus the pressor activity left after treating the primary extracts
with i/io of the volume of 10 N NaOH for 10 minutes at 100° C.
was relatively less influenced by cocaine (Fig. 1) than the untreated
extract.

Effect of ergotamine. The probable relationship between the
unknown pressor substance and sympathomimetic compounds
of the adrenaline type made it desirable to test the action after
ergotamine. A dose sufficient to annul or reverse the pressor ac-
tion of adrenaline (2 mg/kg) strongly decreased the action of puri-
fied preparations, but seldom abolished the pressor action com-
pletely (Fig. 3 and 4).

The obvious difference in action -between adrenaline and spleen
extracts after ergotamine is of interest since it is known from the
investigation on sympathomimetic amines of Baegee and Dale
(1910) that different amines behave differently in this respect.
Thus the methylamino-bases (such as adrenaline) after a suitable
dose of ergotamine produced a pure fall in pressure on the spinal
cat whereas the corresponding amino-base caused a slight rise.
On other occasions the action of nor-adrenaline may be wholly
inhibited or even slightly reversed by ergotamine though never
to the same extent as adrenaline (Stehle and Ellswoeth,
1937).

It was therefore considered that the active pressor extracts
contained a pressor principle of a type differing from adrenaline
and more resembling the type of catechol bases represented by
nor-adrenaline. Catchol-ethanol-amine not being available, a num-
ber of comparative tests were made with a similar substance, dl
3 : 4-dihydroxy-nor-ephedrine (D. N. E.) prepared by the I. 6.
Farbenindustrie and kindly put at my disposal by prof. G. Lilje-
STEAND. This compound was not included in the series investigated
by Baegee and Dale but has been studied by Schatjmann
(1931). The tests with D. N. E. revealed the same type of action
as the amino-bases in Baegee and Dale’s experiments, and ap-
parently D. N. E. resembles nor-adrenaline to a high extent,
quantitatively as well as qualitatively. After a dose of ergotamine
sufficient to reverse the action of adrenaline both the extract
and the D. N. E. retained a slight pressor effect in doses equi-
pressor before ergotamine (Fig. 4).

175

SYMPATHIN E PROPERTIES IN SPLEEN EXTRACTS.

Fig. 5.

A. Blood pressure, cat, chloraloso.

1. 2 fig adrenaline.

2. 0,03 purified spleen extract.

3. 3 fig dihydroxy-nor-ephedrine (D. N. E.).

B. Isolated rabbit’s jejunum.

1. 1 fig adrenaline.

2. 0.01 purified spleen extract.

3. 1 fig D. N. B.

4. 0.5 fig adrenaline.

C. Isolated non-pregnant cat’s uterus.

1. 0.5 fig adrenaline.

2. 0.02 purified spleen extract.

3. 0.05 purified spleen extract.

4. I fig D. N. E.

5. 3 «g D. N. B.

If the difference in action of the spleen pressor substance and
adrenaline after ergotamine can be interpreted as a weaker inhib-
itory action on the blood vessels of the former, one should ex-
pect the extracts to differ in their action from adrenaline also on
other test organs such as intestine or uterus. Equipressor doses
of adrenaline and purified extracts of spleen were accordingly-
tested on these organs.

176

U. B. V. BUIiER.

h. Effect on intestine.

The tests "were made on the isolated rabbit’s jejunum in Tyr-
ode’s solution. A comparison of adrenaline and purified extracts
showed also on this preparation a marked difference, the pressor
extracts causing a smaller inhibitory effect like that of D. N. E.
(Eig. 6). Less purified extracts with depressor action were as a
rule strongly stimulating. On tliis account adrenaline was added
to the purified extract which then produced the usual inhibition,
showing that an effect of this kind was not masked by the pres-
ence of small amounts of stimulating substances left in the
extract.

c. Effect of cat's uterus.

Eurther evidence of a real difference between adrenaline and
the active pressor substance in the spleen extracts was obtained
by testing the active preparations on the virgin or non-pregnant
cat’s uterus. This organ reacts to adrenaline with a pure inhibi-
tion already in small doses, whereas the non-methylated sym-
pathomimetic amines are definitely less active (Barger and
Dale, 1910). The result of the tests was quite decisive, showiag
the usual fall in tone after adrenaline, whereas an equipressor dose
of the pressor extract caused no action or a slight effect only. With
higher doses, again, inhibition ensued which was in good accord
with the action of D. N. E. as shown in Eig. 5. In extracts purified
in different ways varying degrees of inhibitory action have been
observed. Thus of two extracts, in doses having the same effect
on the blood pressure, one being a simple lipid ether extract and
the other further purified by sublimate treatment, the first one
caused a slight inhibition of the uterus, the second no change at
aU, and a dose of adrenaline equipressor with the doses of extracts,
caused the usual strong inhibition. It can thus be stated with a
fair degree of safety that the spleen extracts contain a substance
behaving much like the amino-bases in Barger and Dale’s
experiments.

d. Rabbit’s uterus.

On the rabbit’s uterus the action of spleen extracts like those
of adrenaline and D. N. E. was a purely stimulating one. A com-
parison of equipressor doses revealed, however, that the action
of the spleen extracts was some 5 times smaller than of adrena-

SYMPATHIN E PROPERTIES IN SPLEEN EXTRACTS.

177

Fig. 6. Isolated cat's heart, Langendorff preparation.

1. 1 //g 3 : 4.dihydroxy-nor.ophedrino (D. N. E.).

2. O.012 purified spleen e.\tract._

3. Equipressor dose of (2) purified over amyl alcohol.

4. Twice the dose in (3).

5. 1 jiig adronalino.

line, whereas that of D. N. E. was some 20 times weaker. Even
on this test object, consequently, the spleen extracts showed a
distinct difference against adrenaline.

e. Isolated heart.

Purified extracts of spleen were tested on the isolated perfused
heart of the cat and the rabbit according to Langendorpe and
compared with adrenaline and D. N. E. The latter substances
both stimulated the heart to about the same extent in the same
doses as stated by Schaumann (1931) and the purified spleen
extracts had a very similar effect in equipressor doses (Eig. 6).

/. Effect on the pupil width.

Injections were made either intravenously or intra-arterially
through the common carotid in central direction. A comparison
of the pupil-dilating effect of active extracts and adrenaline
12 — i60215. Acta pliys. Scandinav. Vol. 11.

178

U. S. V. EULER.

Fig. 1. Blood pressure, cat, cUoralose.

1. 0.5 //g adrenaline.

la. Same heated 5 min. at 100° C. -vrith N HjSO,.

2. 0.5 /<g 3 : 4;-dihydroxy-nor-ephedrine (D. N. E.).

2a. Same, treated as in (la).

3. 0.1 purified spleen extract.

3a. Same, treated as in (la).

3b. Same, heated to boiling •with N NaOH.

4. 0.5 /fg D. N, E. treated as in (3b).

5. 0.1 (3) heated 5 min. at 100° C with N NaOH.

6. (After cocaine hydrochloride 10 mg/kg i. m.) 0.1 (3b).

7. 0.1 (3).

8. 0.5 ftg D. N. E.

0. 0.5 /tg adrenaline.

showed a very marked difference in action; adrenaline widened
the pupil much more than equipressor amounts of extracts.
The possibility that impurities with constrictor action were ope-
rating ia the extracts could be ruled out by the fact that addi-
tion of an amount of adrenaline to the extract, less than half of
the equipressor dose, still produced a definitely stronger dila-
tion than the extract. Still a slight widening was regularly ob-
served with the purified extracts, indicating a sympathomimetic
action, yet different from that of adrenahne.

4. Stability, solubility.

The active extracts lost most of their biological activity on
heating to 100° in normal alkaline solution for 5 minutes. When
just heated to boiling after addition of one tenth of the volume
10 N NaOH to the slightly acid lipid-ether extract, about 50 %
of the activity was still left whereas adrenaline or D. N. B. was
completely destroyed. Further purified extracts lost, however.

sniPATHIN E PROPERTIES IN SPLEEN EXTRACTS. 179

as much of their activity as equipressor doses of adrenaline or
D. N. B. l^Tiether the lipid-ether extracts still contained buffering
substances in sufficient amount to prevent a full effect of the
added alkali or whether contaminating substances with protect-
ing action are responsible for the difference cannot be decided.

In the primary extracts (before lipid-ether extraction) a cer-
tain fraction of the pressor activity was left even after 10 minutes’
heating at 100° with one tenth of the volume of 10 N NaOH,
indicating a very stable pressor substance (Fig. 1). As seen in
the figure the remaining effect was not enhanced by cocaine in
the same proportion as the untreated extract indicating the pres-
ence in the primary extracts of a substance resembling tyramine
in its action, which is depressed by cocaine as shown by Taintbr
and Chang (1933), and also resists 10 minutes’ heating at 100°
in normal acid or alkabne solution.

On the other hand there is a certain amount of activity left, still
causing a pressor action after cocaine (Fig. 1 (G)). This remarkable
stability is lost as a result of further purification and may be
related to the presence of stabilizing agents, as in the experiments
by Heard and Eaper (1933) on perfusion of the adrenal gland.

Heating with normal sulphuric acid likewise diminished the
action, though less readily. The follovnng table illustrates the
inactivation by treatment with alkali and acid, denoting per cent
activity left.

,The purified extracts are easily inactivated by treatment with
oxidizing agents such as iodine or potassium permanganate which
IS in accord with the other evidence for catechol bases.

A. Normal alkali
heated to boiling in

Lipid-other

extract

Filtrate from
HgClj-prcoip.

D. N.E.

Adrena-

line

Tyrd-

mine

10 secs

>50

<20

100

5 mins. 100° ....

20—50

100

10 5 100 °

20—30

100

B. Normal acid

6 mins. 100 ° . . .

50—100

100

10 » 100 ° ....

60—100

<20

100

_ The results regarding solubility are probably less significant
since they apply to extracts, containing comparatively large
amounts of impurities in spite of the relative high purification as

180

U. S. V. EULER.

to the biological effect. The solubility in alcohol showed great
variations according to the degree of purification of the extracts.
The solubility in ether or chloroform was negHgible. In con-
formity with this it has not been possible to obtain but a trifling
yield on continuous fluid extraction with ether at pH varying
from 3 to 8.

As already stated above the active substance dissolves to some
extent in an ether solution of organ lipids or pure lecithine.
When other ether soluble vehicles such as triglycerides, oleic acid,
paraffine oil, or bee-wax were used the yield was much less, being
highest for wax. There seems to be a definite quantity of the
pressor substance carried by a given amount of lipids in ether
solution, since the yields have been much the same in each of
20 successive extractions. It has also been possible to show that
adrenaline, added to the extract, is extracted in a similar propor-
tion as the inherent pressor activity.

5, Colour reactions.

Many of the almost colourless purified extracts contained
amounts of pressor activity corresponding to some 50 — 100 jxg
adrenaline per ml, Assumiag a similar chemical composition and
an activity not greatly exceeding that of adrenaline per unit of
weight it should be possible to obtain some of the typical colour
reactions for this substance.

A rather sensitive colour reaction, but of little specificity, is
the Folin-Cannon-Denis test with tungstic acid, which was al-
ways positive but usually much stronger than the reaction
with an equipressor amount of adrenaline probably owing to
impurities.

The simple FeCls test, on the other hand, gave a greenish tint
of the same order of strength as adrenaline for equipressor amounts
in purified extracts, changing to red after alkalization. This
reaction, indicating a catechol nucleus, often proved useful in
estimating the approximate strength of the extracts as to the
pressor activity.

A^Tien the Vulpian iodine reaction was employed according to
the method described by Etjleb (1933), it was not possible, how-
ever, to obtain any red colour, though addition of adrenaline in
amounts less than half of the total pressor activity caused a def-
inite reaction. It was noted, that the extracts obtained a. slight

SYMPATHIN E PROPERTIES IN SPLEEN EXTRACTS. 181

straw yellow colour, sometimes changing to greyish, like that of
dioxiphenyl-ethyl-amine, though this gave an initial strong red
colour with iodine.

From the colour test experiments it may be inferred that the
active substance, probably is a cathechol compound of the same
order of activity as adrenaline. These findings seem to rule out
a number of sympathomimetic amines, such as hydroxytyr-
amine, which is some 50 times weaker in its pressor effect than
adrenaline, and many others. Among the known catechol deriva-
tives only dihydroxy-phenyl-ethanol-amine and its ^-methyl-
substituted derivative seem to come into question on account
of their activity and biological actions.

Of the physical methods for detecting adrenaline that of
Gaddibi and Sohild (1934), using the green fluorescence of adre-
naline in alkaline medium in the presence of oxygen, has the
advantages of being very sensitive and quite specific. The reac-
tion was carried out with equipressor concentrations of purified
spleen extract, adrenaline and dihydroxy-nor-ephedrine, both the
latter in a strength of 1: 100,000. After appropriate dilution to
compare biologically with the adrenaline solution, the purified
spleen extract was almost colourless and showed only a very weak
greenish fluorescence of its own. After addition of alkali the adre-
naline solution showed a strong green fluorescence, whereas no
visible change was noted in the other solutions. Addition of adrena-
line to the spleen extract, corresponding to ^/lo of the pressor
activity (1 in 1 cc) still produced a distinct reaction, showing
that if any adrenaline was present in the extract the amount was
certainly less than the added amount. These results give a def-
inite proof that the active substance is different from adrenaline
as already indicated by the biological tests. At the same time they
give further indirect support to the assumption that the active
substance is identical with or closely related to catechol-ethanol-
amine, which gives a very weak fluorescence reaction, about 2
p. c. of that of adrenaline, as found .by Gaddum and Sohilh.

From these experiments it appears that the active substance
in the purified spleen extracts shows a much closer agreement
with the action of 3 : 4-dihydroxy-nor-ephedrine than with that of
adrenaline. Unfortunately it has not yet been possible to compare
the actions of the extracts directly with 3 : 4-dihydroxy-ethanol-
amine but it is evident from the experimental data in the literature
concerning its biological actions (Barger and Dale, 1910, Greer,

182

U. S. V. EDLER.

Pine;ston, Baxter and Brannon, 1937) that a close relationship
should exist between this substance and the chief active principle
in the extracts.

Comments.

In a preliminary report (Euler, 1915) it has been shown that
considerable amounts of a sympathomimetic ‘pressor’ substance
may he found in a number of fresh organ extracts. Erom the re-
sults put down in the present paper it emerges that the active
substance in spleen extracts, which have been subjected to a
closer study, is related to, but different from adrenaline.

The careful analysis of the action of sympathomimetic sub-
stances on various biological test objects in the classical study of
Barger and Dale (1910) has brought to light certain important
differences between the action of the methylamino-bases of the
catechol group, such as adrenaline, on the one hand, and the ami-
no-bases — and to a lesser degree the ethyl- and propylamino-
bases — on the other. It is possible, and even probable, that the
spleen extracts contain more than one active substance belonging
to the oxyphenyl- or catechol groups but it seems safe to assume
that at least an important fraction of the activity is due to a
substance having the characteristic features of a catechol amino-
base. Direct comparison has shown an extensive agreement with
the isomer of adrenaline, 3 ; 4-dihydroxy-nor-ephedrine which has
been used in lack of the 3 : 4-dihydroxy-ethanol-amine, the non-
methylated adrenaline or nor-adrenaline.

The significance of these findings is closely connected with the
fact that the biological effects of sympathetic stimulation only
partly agrees with that of adrenaline and that a number of sti-
mulation effects show a much closer agreement with the actions
of the amino-bases of the catechol group as pointed out by Barger
and Dale. The following may be quoted from their paper:

“The conception of sympathetic nerve-impulses as acting by
the liberation of adrenine seems to us unsatisfactory for another
reason. It involves the assumption of a stricter parallelism between
the two actions than actually exists. Adrenine has, in common
with the other methylamino-bases of the catechol group, the
property of exaggerating inhibitor as compared with motor
effects. The inhibitor effects of these methylamino-bases are
relatively prominent not only as compared with those of homo-

STMPATHIN E PROPERTIES IN SPLEEN EXTRACTS. 183

logous bases, in particular tbe aminobases, but also in comparison
TOth those of sympathetic nerves. All the sympathetic effects
which are weakly or doubtfully reproduced by adrenine are motor
effect — e. g. pilo-motor action, or contraction of the trigone
of the cat’s bladder. On the other hand certain inhibitor effects,
such as inhibition of the fundus of the cat’s bladder, or of the non-
pregnant uterus of the same animal, are more easily and complete-
ly produced by adrenine than by nerve-stimulation. Similarly
certain normally motor effects of adrenine are reversed by smaller
doses of ergotoxine than are needca for the reversal of the corre-
sponding motor effects of stimulating sympathetic nerves. In these
respects the action of some of the other bases, particularly of the
amino- and ethylamino-bases of the catechol group, corresponds
more closely vdth that of symphatetic nerves than does that of
adrenine. To suppose that such bases and sympathetic nerve-
impulses alike owe their action to the liberation of adrenine seems
to us to create additional difficulties of conception.”

Barger and Dale’s inferences regarding the nature of a pos-
sible mediator substance involved in sympathetic stimulation
(adrenergic fibres) has received strong indirect support from the
experimental work of Cannon and his associates, Bacq and
Rosenblueth, on sympathin. They showed that stimulation of
certain sympathetic nerves produced remote actions, proving the
liberation of a kind of substance, which was obviously related
to but, in certain respects, different from adrenaline. The latter
substance was, however, regarded as the primary mediator.
By assuming the formation of special cell reaction products,
sympathin E and I, Cannon and Rosenblueth (1933) made an
attempt to interpret their observations. Neither of the postulated
sympathins has, however, hitherto been isolated or even prepared
in such a way that their properties could be put to a decisive test.

The similarity between the excitatory effects of sympathetic
stimulation and of non-methylated adrenaline, as originally shown
in Barger and Dale’s experiments, led Bacq (1934:) to express
the thought that the latter was the excitatory mediating substance
and adrenaline the mediator where inhibitory actions were pre-
vailing. This view has been shared by Stehle and Ellsworth
(1937). Similarly Greer, Pinkston, Baxter and Brannon
(1936) point out, as a result of their experiments, that stirnulation
of the hepatic nerves causes effects which are in better accord
with the action of nor-adrenaline than with adrenaline. The pres-

184

u. 6. V. eui,>:k.

ont findings have given direct proof of tlie occurrence in spleen
extracts of a substance wth similar properties as those charac-
teristic of syinpathin E though it should he noted that it also has
a weak inhibitory action. Since the EeClj catechol reactions for
equipressor solutions of adrenaline and spleen extract are similar
in strength it is inferred that the active substance must be of the
same order of activity os adrenaline. This gives some indication
as to the exact nature of the unknowm substance, since a similar
quantitative relation so far is known to exist only for adrenaline,
nor-adrenalinc and dihydroxy-nor-ephcdrinc. Assuming that the
active substance in spleen extracts is also the substance liberated
on sympathetic stimulation in this organ the probability of the
identity of the latter with nor-adrenalinc is greatly increased

Snmmarj'.

1. Extracts of fresh cattle spleen possess a pressor activity
equivalent to some 10 //g adrenaline per g of ti.ssuc.

2. The purified substance increases the heart rate and raises
the blood pressure of the cat in chloralosc anaesthesia.

3. Tlic pressor action is enhanced by cocaine.

4. Ergotamine in doses which annul or reverse the pressor
action of adrenaline is less active in depressing the action of
purified spleen extracts, which in this respect resembles certain
catechol amino-bases, such as nor-ndrcnaline or 3 ; 4-dihydroxy-
nor-ephedrine (D. N. E.).

5. Adrenaline inhibits the isolated rabbit’s intestine and the
non-pregnant cat’.s uterus more powerfully than equipressor doses
of spleen extracts or D. N. E.

G. Purified spleen extracts, like D. Is. E., are less active in
stimulating the rabbit's uterus than equipressor doses of adrena-
line.

7. Purified spleen extracts and D. N. E. have a weaker pupil
dilating action than equipressor doses of adrenaline.

8. Purified spleen extracts stimulate the isolated heart in
much the same way as equipressor doses of adrenaline and
D. N. E.

9. Purified spleen extracts and D. If. E. do not give the fluor-
escence reaction characteristic of adrenaline in equipressor con-
centrations.

EYMPATIirN E PIVOPERTXIS IN SPLEEN EXTRACTS. lb»

10. Purified spleen extracts and D. N. E. give tlic EeClj colour
reaction to about tbe same strength as equipressor concentrations
of adrenaline.

11. ' The biological tests, colour and fluorescence reactions of
purified spleen extracts thus bear a good resemblance to those of
nor-adrenaline or D. N. E. and differ from those of adrenaline.

12. The similarity between the action of the purified spleen
extracts and the postulated sympathin E on the one hand and
nor-adrenaline or D. N. E. on the other is pointed out.

This work has been supported bj- a grant from Astra Ltd,
Sodertiilje, for which I wish to express my sincere thanks.

Roforoncos.

Bacq, Z. ]\I., Ann. Physiol. 1934. 10. 407,

Barger, G., and H. 11. Dale, .T. Physiol. 1910. 41. 19.

BrxGEL, A., and E. Strauss, Dtsch. Arch. klin. Med. 1909, 96. 176,
Canxox, W, B., and K. LissAk, Amer. .1. Physiol. 1939. 125. 765.
Canxon, AV. B., and A. Rosenblueth, Autonomic Neuro-Effector Sys*
stems, New York 1937.

CoLLiP, J. B., J. Physiol. 1928. 66. 41 6.

Collip, j. B., Trans. Roy. Soc. Can. 1929. 22.

Enoer, R., Dtsch. Arch. klin. Med. 1942. 1S9, 75.

Euler, U. S. v., Biochem. Z. 1933. 260. 18.

Euler, U. S. v., J. Physiol. 1934. SI. 102.

Euler, U. S. v.. Nature 1945. 156. 18.

Euler, U. S. v., and C. 6. Sgilmiteulow, Acta Physiol. Scand. 1944.
8. 122.

Euler, U. S. v., and T. S.iostrand, Naturwiss. 1943. 31. 145.
Frohlich, a., and 0. Loewi, Arch. e.xp. Path. Phnrmnk, 1910. 62. 159.
Gaddum, j. H., and M. A. Kh.vyyal, cited by J. H. Gaddum in “Gc-
fassorweiternde Stoffc der Gewebe", Leipzig 1936.

Gaddum, J. H., and H. Schild, J. Phy.siol. 1934 . 80. 9 P.

Grade, F., 0. Krayer, and K. Skelkopf, Klin. AVsebr. 1934. 13. 1381.
Greer, C. M., J. 0. Pinkston, J. H. Baxter, and E. S. Bran.non,
J. Pharmacol. 1937. 60. 108.

Hartwich, a., and G. Hessel, Zbl. inn. Med. 1932. 53. 612.

Heard, R. D., and H. S. Raper, Biochem. J. 1933. 27. 36.

James, A. A., N. B. Laughton, aud il. B. Macallum, Amer. J. Physiol.
1926. 75. 392.

LissAk, K., Amer. J. Physiol. 1939. 125. 778.

Loewi, 0., Pfliig. Arch. ges. Phy.siol. 1921. 189. 239.

Loewi, 0., Ibidem. 1936. 237. 504.

Oliver, G., and E, A. Schafer, J. Physiol. 1895. 18. 230.

186

U. S. V. EULEn.

Raab, W., Biochcra. J. 1913, 37 . 470.

Rooek, H., Pr. Med. 1922. 30 . 441.

SoiiAXJMANN, 0., Arch. exp. Path. Pliarmak. 1931. 160 . 127.
Steele, B. L., and H. C. Ellswokth, J. Pharmacol. 1937. 59 . 114.
Tainter, M. L., and D. K. Ciiako, Ibidem 1920. 30 . 193.

Vincent, S., and W. Sheen, J. Physiol. 1903. 29 . 242.

WiLLiAJis, J. R., and E. B. GROSS>rANN, Amcr. J. Physiol. 1938. 123 .
364.

Biological Institute, Garlsbcrg Foundation, Goponbagon,

Protein Metabolism of Tissue Cells in vitro. 4.

The Properties of Malt Extracts and of Glutathione
as Accessory Growth Suhstances.

TAGE ASTRUP and ALBERT FISCHER.

Received 28 November 1915.

In continuation of our previous investigations on substances
interfering vrith the protein metabolism of tissue cells in vitro,
(Fischer 1941, 1942), we have recently studied the accessory
growth substances contained in yeast and barley malt (ibsxRur,
Fischer and Volkert 1945, Astrup and Fischer 1945). In this
paper we are presenting further investigations on the properties
of the accessory growth substances in barley malt together with
investigations on glutathione. The technique is the same as
described in previous papers, dialyzed media being used.

Experimental.

1. Crude Extract. Method II.

For the easy preparation of larger amounts of malt extract the
method described previously (Astrup and Fischer 1945), was
found inconvenient. This method 'ivill bo referred to as Method I
for the preparation of crude extract. The following method II,
in which the time consuming centrifugation of the mixture of
water and malt was avoided, was therefore used.

200 g of pulverized barley malt is added to 500 ml water at 35—40°
and stirred at this temperature for one hour, l.c liter of ethanol
(96 per cent) is added, and after standing for a couple of days, the
mixture is filtered on a Btichner funnel. The filtrate is evaporated in

188

TAGE ASTRUP AKD ALBERT FISCHER.

vacuo on a vatcrbatli (<. 50°) to about 300 ml and the lipoids removed
by filtration. After dilution to tbe original volume and neutralization,
it is sterilized and tested on tlic tissue cultures in tbe usual manner.

This malt extract contains about the usual amount of nitrogen
(0.8 — 1.00 mg per ml). By the addition of ethanol considerable amounts
of lipoids and coloured matter are extracted
from the malt, but these are subsequently re-
moved by the evaporation and filtration of
the aqueous solution.

The extract seems just as active as an ex-
tract prepared in the usual manner, see
fig. 1. In the following experiments the crude
extract was as a rule prepared after tliis
method and eventually concentrated further
in vacuo to one tenth of the original volume.
In this form it could be kept several months
in the ice box without deterioration.

2. Glutathione.

As glutathione is a universal constituent
of living cells and is an activator for cer-
. . . , . tain enzymatic reactions in the organism,

extract, Jfethod II (V- f- 1- proteoIytic processes, it might be of in-
243.1). Culture no. S4S0, terest to test this substance. In previous

Culture no. 5479,110 addi- papers (FiSCHEB and AsTRur 1942, Astrup
tio"- Fischer and Volkert 1945), glutathione

was investigated after autoclaving, . as it

Fig. 2. Action of glutathione. Culture no. 8347, 0.2 ml of purified malt extract
(V.231.1, corresponding to V-217.6 by Fischer and Astrup (1945)). Culture
no. 8348, 0.2 ml of 0.2 per cent glutathione (V-241). Culture no. 8573, no addi-
tion. Culture no. 8574, 0.2 ml 0.1 per cent glutathione (V-248.1).

/ 2 ■ J 4 S a 7 f 2 3 4 5 7

Fig. 3. Photographs of living heart fibroblasts in Carrel flaslcs. hlagnif. 40 X .
No. 8949, Mother liquor from malt extract treated with cadmium lactate after
Binet and TfeKer (V-250,1). No. S950, glutathione (V-252,3). No. S927, oxydized

glutathione (V-259).

190

TAGE ASTRUP AKD ALBERT FISCHER.

should be compared Trith firactions active after such treatment.
We now tried the substance after sterile filtration. Glutathione
was prepared from bakers yeast after the methods of Pirie
(1930) or Schboeder, Collier and Woodward (1939). Solutions
were prepared in physiological saline, almost neutralized and
sterile filtered through sintered glass filters.

Glutathione, when used in not too low concentrations, shows
some activity on the tissue cultures. Comparatively large areas
of growth may be obtained, but the cells are not of normal appear-
ance and very granulated; the cells disintegrate in the course
of 4 — 6 days. Its action is rather similar to the action of cystine,
though larger areas may be obtained. Some results are shown
in Kg. 2 and 3. Oxidized glutathione shows the same activity.
Autoclaving does not completely destroy it.

The experiments with glutathione show that this substance,
although active to some extent, is not responsible for the full
effect of the malt extracts. Another question is, to what extent
it ccmtributes to this effect, or if other active substances are the
sole cause. This may be disclosed by preparing glutathione-free
fractions of the malt extract. The content of glutathione in differ-
ent extracts was therefore determined after the colorimetric
method of Fujita and Humata (1939).

3. Glutathione in the Active Extracts.

Yeast is a rich source of glutathione, and the active yeast
extracts previously investigated (Astrup, Fischer and Yolkert
1945) proved to contain amounts of this substance, sufficient to
allow its isolation over the cuprous merkaptide compound. Malt
extract also contains glutathione, but in considerably smaller
amounts. Prepared after Method, I it contained about 0.14 mg
total glutathione per ml; of this 0.08 mg per ml was in the reduced
form. These amounts are far smaller than the concentrations of
glutathione necessary in pure solutions for producing growth
of any extent (about 2 mg per ml), and do not indicate any signif-
icance of glutathione for the growth promoting properties of
malt extracts.

It was attempted to remove the glutathione ftom the solutions
by means of Cu,0 (HoPKms 1929) or using the cadmium method
of Biket and Weller (1935), but the resulting solutions still
contained small amounts of glutathione.

PKOXEIN METABOLISM OF TISSUE CELLS IN VITRO. 191

Y-245,1 is prepared by Method I and purified by one precipitation
•with ethanol at basic reaction. It contains 0.54 mg N per ml and 0.14
mg total glutathione per ml. After concentration to Vj volume a small
precipitate separates after treatment with CujO in the usual manner.
Copper is removed from the precipitate, and it is tested on tissue cul-
tures in four times the concentration of the original extract. It con-
■tains now O.ii mg total glutathione per ml and O.og mg K per ml
(V-245,2), but shows no effect.

Cn,0 precipitation, V-245,3 (0.2 ml). Culture no. 8630 original malt extract,

V;245,l (0.2 ml).

HsS 04 is removed from the centrifugate from the CujO-treatment
by means of Ba(OH)s, and diluted to the original volume (V-245,3).
It contains O.oo mg total glutathione and O.io mg N per ml and seems
just as active as the original solution V-245,1, see fig. 4. Similar results
were obtained after treatment with cadmium (V-260) and seem to
exclude any major significance of glutathione for the growth pro-
moting properties of malt extracts.

Some malt extracts were treated •with HjS in order to convert
all glutathione present into the reduced form to facilitate the
complete precipitation of it. While this object was not reached,
it was found, that during the treatment with HoS a considerable
amount of a yellow precipitate separated. This treatment as a
rule did not reduce the actmty, but considerable amounts of
nitrogen were removed from the solution.

Neither treatment •with Hj or COj influences the aoti^vity nor

X92 '-TAGE astrop and albert fischbr.

produces R pfecipitcite. A-t Deutral or alkulino reRction no pre-
cipitRte is formed by means of H,S. The nRture of the precipitate
is unknown and was not investigated.

4. The Sugar Content of the Malt Extract.

Many of the difficulties met with in the purification of the
extracts found to be due to the considerable amounts of sugar
contained in the solution. The sugar content was therefore followed
during the operations by means of the method of HIagedorn and
Jensen. Most of the sugar present in the extracts must be assumed
to be maltose, of which one mol. (342) requires about fifty per
cent more oxygen than one mol. of glucose (180), see Sobotka
and Eeiner (1930) and Micheel (1939). This must be taken into
consideration, when making use of the amoimts of sugar generally
determined in the Hagedorn and Jensen method, and which are
stated ^vithout conversion in the following investigations. Also
glutathione influences the sugar determination (B!erbert, BoxTRI^e
and Groen, 1930), but this is of no importance in this connection,
due to the small amounts of glutathione present.

A crude extract II (V-273,1) showed a content of 26. i mg sugar per
ml solution before the alkaline treatment and 19.9 mg per ml after.
After treatment with HjS (V-273,2) it contained 16.5 mg sugar per ml.
Eone of these operations thus removed any considerable part of the
sugar content.

A concentrated malt extract (V-275) after alkaline precipitation and
treatment with HoS contained 21.3 mg sugar per ml (diluted to the
original volume) (V-275, 2). After precipitation with acetic acid and
ethyl alcohol, see Astrot and Fischer (1945), the active solution
contained 10. o mg sugar per ml (V-275, 3). This treatment thus removed .
about half of the amount of sugar present.

Several attempts were made to remove by simple chemical means
the large amounts of sugar stUi present but without much success.
Fractionated precipitation with methyl alcohol was tried. A concen-
trate made acid to congo paper with diluted HCl was soluble in methyl
alcolhol. After the addition of absolute alcohol, an inactive precipitate
containing a considerable amount of sugar appeared, but this treat-
ment seemed not more advantageous than the precipitation with acetic
acid and ethyl alcohol, (V-273, V-274).

In previous investigations on the accessory groivth substances pres-
ent in yeast (Astrup, Fischer and Volkert 1945), we found,, that
treatment with actis'e carbon removed a considerable part of nitrogen
containing impurities and only small amounts of the active substauces.
The partially purified (acetic acid) malt extracts were treated with active
carbon in order to find means for absorbing the active substances.

PROTEIN METABOLISM OP TISSUE CELLS IN VITRO. 193

but also ia this case they were to a large extent left behind in the
solution. Only minor amounts of sugar and nitrogen were removed at
neutral reaction by one or two treatments with active carbon, and the
resulting solution was almost as active as the original solution (V-273,7,
V-276,5). Repeated treatment (five times) removed however the active
substance and about half the amount of sugar and nitrogen (V-273, 9).
In acetic acid solution three successive treatments with active carbon
removed most of the active substances, only small amounts of sugar
and considerable amounts of nitrogen {V-275,5). But it was not found
possible to recover the active substances adsorbed on the carbon.
Treatment with 33 and 50 per cent acetic acid, NaOH to reaction on
phenolphtalein and diluted NH 4 OH yielded no results. Treatment of
less purified solutions, containing more sugar, with active carbon
resulted in still poorer adsorption of the active substances (V-277,2).

Discussion.

Glutathione was found by Baker (1928) to increase the growth
of fibroblasts in vitro. Hueper and Russell (1933) stated that
concentrated embryo extracts as used in tissue culture contained
40 mg glutathione per 100 ml and found cysteine and glutathione
to increase cellular growth. In plasma glutathione disappears
rapidly. The growth promoting substances extracted from sprout-
ing maize were assumed not to be glutathione, even if SH-groups
eventually played a role (Paulmann 1939),

According to our investigations glutathione has definite effect
on dialyzed media, but is not able to replace an extract of
barley malt. It is also generally accepted that glutathione rapidly
disappears in plasma (Oberst 1935, Woodward 1939), and that
the content of glutathione in blood is exclusively due to the
glutathione contained in the red blood cells, cf. UssiNG (1943).
It may therefore also be assumed, that the accessory growth
properties of native plasma or plasma dialysates are not due to
a content of glutathione, but according to later experiments by
Numata (1940) plasma still contains small amounts of glutathione
in the oxidized form, about 7 mg per 100 ml. This is of the same
order as found by us in the active malt extracts, but is consider-
ably smaller than the amounts found necessary to produce any
significant growth of the tissue cultures on the dialyzed media.

As the malt extracts do not contain any considerable amoimts
of glutathione, and preparations containing stiU lesser amounts
are fully active on the tissue cultures, this indicates that gluta-
thione is of only minor importance, if any, for the accessory
13 — i6021S. Acta phys. Bcandinav. Vol. 11.

194

TAGE ASTEOP AND ALBERT FISCHER.

growth, properties of malt extracts. By the methods investigated,
it was not possible quantitatively to remove glutathione from
the solutions; probably the large content of sugar was responsible
for this.

Other known substances present in living cells could probably
be identical with the substances sought for, either alone or in
conjunction with glutathione. We have tried in this manner
ascorbic acid and panthothenic acid, but without any success.
A solution of 0.2 per cent ascorbic acid (after almost neutrahza-
tion and sterile filtration) was inactive with or without the addi-
tion of glutathione (V-258). This was also the case with pantho-
thenic acid (0.2 per cent solution, sample from “Hoffmann-La
Koche”, V-238, V-294), although experiments by Pratt and
Williams (1939) indicate a probable action on the respiration of
tissues in vitro. Several of the properties of our substances seem
similar to the properties of panthothenic acid, cf. Williams,
Truesdail, Weinstock, Rohrmann, Lyman and McBurney
(1938). In addition to the chemical difficulties met with in hand-
ling such substances, we meet, however, also the obstacle of being
without any means for quantitative determination of the amount
of active substance present, as the tissue culture method only
yield a rough idea about the potency of the preparations. The
only result of any reliability is obtained, when the solutions are
practically inactive.

Mangano chloride was tried, with and without the addition of
glutathione, but was found inactive, p-amino-benzoic acid was
inactive and the same was the case with glycocoll.

Some of the properties disclosed in addition to those previously
mentioned (Astrup and Fischer 1945), are the following; A re-
lative stability against diluted aqueous and alcohohc alkah in the
cold, but not when heated; a higher solubility in methyl alcohol
than in ethyl alcohol. The active substance is not adsorbed on
barium sulfate or franconite and only with difficulty on active
carbon. Treatment with H^S does not destroy the substances.
By fractionation it follows as a rule the sugar fraction, and it was
not found possible by simple chemical means to remove the sugar
from the solutions without interfering with the active substances.
Later it was found, that a fermentation with yeast removed
nearly all sugar without introducing new impurities nor affecting
the active substances, and the purification based on this treat-
ment is now under investigation.

PROTEIN METABOLISM OF TISSUE CELLS IN VITRO.

195

Concerning the function of the active substances, it may be
mentioned that in reality we do not know in which processes of
the living cells they interfere. Wc have assumed, that it is a ques-
tion of the protein metabolism, and that this is concerned seems
unquestionable, but we do not know, if the substances interfere
directly with these reactions, or if they are necessary for the living
cells in other respects, a normal protein metabolism being only
a sign of normally functioning cells.

Suniinary.

1. It is found, that glutathione is an accessory growth sub-
stance for tissue cultures, but that it does not suffice to give
the cultures normal growth and appearance.

2. The malt extracts previously investigated contain small
amounts of glutathione, but this substance is not responsible
for the complementary effect of the malt extracts on dialyzed
media.

3. Further properties of the active fractions arc described.

Most of the analjd;ical determinations in this work were carried
out by stud, polyt. Karen Buhl Christensen. For valuable
facilities in the preparations our thanks arc due to AjS “Ferrosaii",
Oopenhagen, and the barley malt was placed at our disposal by
the Carlsberg Breweries.

Litorataro.

Astrup, T., and A. Fischer, Acta Physiol, Scand. 1915. 9. 183.
Astrup, T., a. Fischer, and M. Volkert, Ibidem 1915. 9, 134.
Baker, L. E., Science 1928, GS. 459.

Binet, L., and G. Weller, Bull. Soc. Chcra. Biol. 1935. 16. 1285.
Fischer, A., Acta Physiol. Scand. 1941. 2. 143.

Fischer, A., Naturwisscnschnften 1942. 30. 6C5.

Fischer, A., and T. Astrup, Pflug. Arch. ges. Physiol. 1912. 245. 633.
Fujita, a., and I. Numata, Biochem. Z. 1939. 300. 246. 257.
Herbert, F. K., M. C. Bourne, and J, Groen, Biochem. J. 1930
24. 291.

Hopkins, F. G., J. biol. Chem. 1929. 84. 269.

Hueper, W. C., and M. A. Russell, Arch. exp. Zellforsch. 1933. 14
483. ’

Micheel, F., Chemie der Zucker und Polysaccharide. 1939. p 60
Nusiata, L., Biochem. Z. 1940. 304. 404.

196

TAGE ASTRDE AND ALBERT FISCHER.

Obeest, F. W., J. biol. Chem. 1935. 111. 9.

Paulmann, F. K., Biochem. Z. 1939. 800, 153.

PiEiE, N. W., Biochem. J. 1930. 24. 51.

Pratt, E. F., and E. J. Williams, J. gen. Physiol. 1939. 22. 637.
SoHROEDER, E. F., V. CoLLiEE, and 6. E. Woodward, Biochem. J.
1939. 33. 1180.

SoBOTKA, H., and M. Reiner, Ibidem 1930. 24. 394.

UssiNG, H. H., Acta Physiol. Scand. 1943. 5. 335.

Williams, R. J., J. H. Truesdail, H. H. Weinstock, E. Rohemann,
C. M. Lyman and C. H. McBurney, J. Amer. Chem. Soc. 1938.
60. 2719.

Woodward, 6. E., Biochem. J. 1939. 38. 1170,

From the Laboratory for the Theory of Gymnastics, University - of

Copenhagen.

Aerolbic Recovery after AnaeroMosis
ill Rest and Work.

ERLING- ASMUSSEN.

Received 9 December 1945.

The energy required for maintenance and functioning of the
organism is ultimately derived from oxydations of the foodstuffs.
In a steady state of rest or work the oxygen intake, therefore,
is an expression of the energy production. Under certain condi-
tions, however, the energy requirements are greater than the
instantaneous oxydations can respond to; the energy then will
he liberated by anaerobic processes and an oxygen debt is con-
tracted. In subsequent periods of revovery this oxygen debt will
be “repaid” as shown by the excess oxygen intake folloiving a
period of anaerobic activity. It is the purpose of this paper to
investigate the efficiency of this repayment.

The anaerobic conditions studied were partly experimental,
viz. produced by blocking of the circulation to the lower extrem-
ities, partly those prevalent during the initial stages of physical
exercise.

Methods.

Oxygen uptakes were determined by the DoUGLAS-bag method,
the air analyses being done on the Krogh-Haldane apparatus. Blood
lactates were determined after Edwards (1938) on finger blood.

A blocking of the circulation to the legs was effected by means of
pneumatic cuffs round the thighs as proximally as possible. On sudden
release of the pressure in the cuffs a fall in the arterial blood pressure
will occur which in turn will be followed by a compensatory increase
in circulation rate, emptying of blood depots etc. and an increase in

198

ERtilNQ ASMUSSEK.

oxygen uptake wMch, kowever is largely used to oxygenate the venous
blood of the blood depots. (Comp. Asmussex, Christensen and Niel-
sen (1938, 1939)) In order to avoid this, the subject was placed with
the legs in a slightly elevated position (Asmtjssen, Christensen and
Nielsen (1939)) and the pressure in the cuffs was released slowly so
that no sudden fall of blood pressure occurred.

The work experiments were performed on a KROOH-bicycle ergometer
which for these experiments had been furnished with a reclining chair
instead of the ordinary saddle, so that the subject could pass from
rest to work and vice versa with a minimum of extra work. The chair
was adjusted so, that also during exercise the extra work necessary for
fixating the body was practically nil, the arras, head and back of the
subject resting relaxed on proper supports.

All the experiments were made under standard conditions in the
morning. A resting period of 1 hour in the lying or recumbent position
preceded each experiment. As subject served E. A., a normal male,
38 years old, height 172 cm, weight 70 kg, well accustomed to and
trained in all the experimental procedures.

Besults.

a. Circulation blooJced in rest.

In these experiments the subject was lying on his back •with
the legs slightly elevated. When the resting oxygen uptake of the
subject had been determined in two 6 minutes’ periods the cuffs
round his thighs were inflated, and one or more determinations
of the oxygen uptake were made. In confirmation of earlier ex-
periments (Asjiussen, Christensen and Nielsen 1939, b) it
was found to be about 20 cc/min lower with the circulation to the
legs blocked than in the normal condition. (Mean of 18 experi-
ments; 19 ± 1.3 cc/min). After 5 to 23 minutes the circulation
to the legs was restored, and the oxygen taken up during recovery
was determined quantitatively until resting values again were
found, usually after less than 8 minutes. The oxygen was calcula-
ted and plotted against the time of anaerobiosis as shown in fig. 1.
In the same figure is shown the magnitude of the oxygen deficit,
calculated from the assumption that this is 20 cc/min. It is evident
from the figure that when the circulation to the legs has been
blocked for more than 5 minutes, the excess oxygen uptake during
recovery is greater than the debt contracted during the anaerobic
period, and the more so the longer the period of occlusion has
been.

Determinations of the blood lactates during recovery showed
a slight increase of 4 to 5 mg pCt in the experiments with the

ABKOBIC EEGOVERY AFTER ANAEROBIOSIS. 199

Fig. 1. Subject at rest. Full line and points: Excess oxygen taken up after resto-
ration of circulation to the legs. Dashed line: Oxygen deficit caused by the occlu-
sion of the circulation to the legs. Abscissa: Duration of occlusion in min.

longest time of occlusion, wliereas the blood lactates "were unal-
tered in the experiments of shorter duration.

It might be objected' against this result, that perhaps tbe muscles
of the thighs have contracted involuntarily during the period of
blocking, thus making the oxygen debt greater. Actually the
subject -was lying completely relaxed, and tests in wliicb, an
Adriax-Beon'K needle electrode, connected through an amplifier
■with a loudspeaker -was thrust into the muscles of the leg, made
it clear that no muscular movements occurred.

Judging from fig. 1 it must be said that tbe organism must pay
a rather heavy “interest” — up to about 90 pCt — ■ on the “debt”
it contracted during the time of occlusion. The ratio deficit:
repayment is 1 after 5 min but decreases to 0.54 after 23 minutes
of blocked circulation.

i. Oirculation blocked during exercise-
These experiments were performed 'while the subject was work-
ing on the bicycle ergometer at a rate of 230 nokg/min. One series
was made at the start of the work and lasted 3 minutes. The total
oxygen requirement of the work was determined both in tbe normal
condition and with the circulation to the legs blocked previous
to work. In the latter case the circulation was restored imme-
diately after cessation of work.

200

ERUNG ASMGSSEN.

Table 1.

Normal start

Start ■with ciroul. to legs blocked

Date

Excess Oj
during work
liters

Excess O 5
in recovery
liters

Total O 2
requirement
liters

Net

efficiency

pCt.

Date

Excess'_0,
during work
liters

E.xcess Oj
in recovery
liters

Total O 2
requirements
liters

Net

efficiency 1
pCt. 1

“A • ■

1.53

0.55

2.08

16.2

“A ••

0.84

1.75

2.59

13.0

1.54

0.45

1.99

16.9

«A.-

0.85

1.42

2.27

14.8

”A • .

1.40

0.49

1.95

17.3

='A--

0.87

1.73

2.00

12.9

“A . .

1 . 6 C

0.34

2.00

16.8

V«--

1.03

1.64

2.67

12.6

"-A . ■

1.5G

0.73

2.29

14.7

’A--

0.94

1.46

2.40

14.0

*A ••

1.52

0.57

2.09

16.1

”A

0.96

1.32

2.28

14.8

mean

1.55

0.52

2.07

16.3

mean

0.92

1.55

2.47

13.7

The results of these experiments are tabulated and averaged
in table 1. The table shows that the total oxygen requirement for
performing 3 X 230 mkg of work is considerably greater when
part of the muscles have been working in an anaerobic condition
than when they perform the work under ordinary conditions.
The net efficiency of the work is 16.3 pCt in the normal condition
but only 13.7 pCt when part of the work was done anaerobically.
The deficit due to the blocking of the circulation must be 1.55 —
— 0.92 =0.63 liters of oxygen, and the extra oxygen taken up
after the restoration of the circulation is on an average 1.55 — 0.52
= 1.03 liters of Oj.

In another series of experiments the blocking of the circulation
to the legs was performed in the steady state of work. The work
went on with blocked circulation at the same rate for 3 minutes,
after which — during uninterrupted work — the circulation was
restored. The work was now continued until the normal steady
state value of oxygen-uptake was reached again, usually after
about 10 minutes. The results are shown in table 2.

Table 2 shows that the oxygen uptake — less the resting con-
sumption — in the steady state is on an average 0.56 1/min and that
it decreases to 0.38 1/miu when the legs are cut off from the cir-
culation. 3 minutes of work in this condition give an oxygen
deficit of 0.55 liters, but the oxygen uptake in excess of the steady
state consumption during the subsequent recovery is 1.36 liters.
The efficiency of the work if done first anaerobically, therefore,
0.55

is only or on an average 43 pCt of the efficiency of work

1.36

when this is done aerobically.

AEROBIC RECOVERY AFIER ANAEROBIOSIS.

201

Table 2.

Date

Kxcess Oj
steady state
liters/min

Excess Oj
blocked circul.
liters/min

Oj-deficit in

3 min. block,
liters

Excess Oj
during recovery
from block, liters

Net efficiency
steady state
pCt

Not efficiency
of “anaerobic”
work pCt.

’V.

0.55

0.35

0.60

1.37

20.5

9.0

“A

0.59

0.39

0.60

1.57

19.1

7.3

0.50

0.38

0.54

1.27

20.1

8.6

«A

0.51

0.37

0.42

1.85

22.1

5.1

“A

0.58

0.39

0.57

1.01

19.4

10.9

“A

0.58

0.39

0.57

1.08

19.4

10.5

mean

0.5G

0.38

0.55

1.36

20.1

8.2

As the net efficiency of this work under aerobic conditions is
20.1 pCt, and assuming, the aerobic efficiency of the muscles
that can be cut off from the circulation to be the same, one comes
to the conclusion, that the efficiency of this work performed
anaerobically ivith a delayed aerobic recovery is only 8 to 10 pCt.

In earlier experiments (Asmussen, Christensen and Niel-
sen (1938), not published) the corresponding ratio, “anaerobic"
efjtciency: “aerobic" efficiency, was found to be: 0.57 in 5 minutes
of work at 180 mkg/min (2 experiments), 0.60 in 2.5 minutes at
360 mkg/min (5 experiments) and 0.71 in 1.2 minutes work at
720 mkg/min (1 experiment).

Determinations of the blood lactates after restoration of the
circulation showed an increase in 1 to 3 minutes to values of 35
to 40 mg pCt, then a gradual decrease to normal values after 10
to 15 minutes.

c. The initial stages of work.

It is generally assumed, that the initial stages of work in which
the oxygen intake does not cover the oxygen requirement, are
performed partly anaerobically. Under such circumstances, and
assuming the low efficiency of aerobic recovery after anaerobiosis
found in the above experiments to be characteristic for anaerobic
work, one might expect to find a lower efficiency or a higher
oxygen requirement during the first minutes of work than in the
steady state, where Oa-requirement and Oj-uptake are identical.

In order to test this assumption a series of experiments was
performed in which the subject worked at rates ranging from

202

EELINQ ASMDSSEK.

Fig. 2. Oxygen requirement for 3 minutes of work of varying intensity.

• • Oxygen requirement computed from oxygen uptake of 3 min. at start,

plus recovery.

X X Oxygen requirement computed from oxygen uptake in the steady state,

q 1- Oxygen deficit at start.

O O Oxygen debt repaid in recovery

• • Blood lactates after 3 min work

X X Blood lactates in the steady state

230 mkg/min to 1150 iixkg/in,in. The total oxygen uptake of work
and recovery — less the resting consumption — of the 3 first
minutes of work was compared with the excess oxygen uptake of
3 minutes of work in the steady state. 2 to 6 experiments at each
grade of work were made. The results are averaged and presented
in fig. 2, which shows that at all grades of work except that of
920 mkg/min, the oxygen requirement for 3 minutes of work
is greater in the initial 3 nainutes of work than in 3 minutes of
steady state. The different result at 920 mkg/min, where they are
equal, was verified by repeated experiments, and no plausible
explanation for this deviation from the rule can be offered. In
fig. 2 the oxygen deficit at the start of work and the oxygen debt
repaid during recovery are also presented. They show as might
be expected that the differences in oxygen requirements of the
initial stage and the steady state is due to the fact that the oxygen

AEROBIC RECOVERY AFTER ANAEROBIOSIS.

203

•dett repaid in recovery is greater than the oxygen debt contracted
at the start of work. The two upper curves show the blood lacta-
tes 1) at the end of the 3 minutes of work in the start, and 2)
in the steady state of work, i. e. after about 16 minutes of work.
Any direct correlation of the blood lactates to either the oxygen
debt or the difference between oxygen debt repaid and oxygen
deficit at start is not disclosed.

There are certain sources of error, that might tend to make the
differences found in these experiments too great. We have care-
fully tried to avoid them and shall list them here in order to show,
that the results presented really, as we believe, depict the actual
facts.

1. Active movements of the subject in order to attain the work-
ing position will tend to increase the initial O.-uptake, thus mak-
ing the calculated Oj-deficit too small. 2. Corresponding move-
ments after cessation of work will increase the amount of Oj
taken up dming recovery. The combined effects of 1 and 2 will
make the total Oj-uptake too high. We consequently have avoided
all active movements, that not directly belonged to the work.
3. In the steady state the muscles have been “warmed-up” and,
therefore, according to experiments by Asmussen and Boje
(19f5), are able to perform maximal work more efficiently. This
would make the Oj-uptake in the steady state lower. We tried
to heat the muscles of thighs and buttocks by short wave diathermy
previous to the start of the 3 minutes initial work, but found
no effect whatever on the amount of oxygen required for this
small work. 4. In the first seconds of work the fly-wheel of the
ergometer must be accelerated up to the speed used in the work.
This will add to the work performed in the initial stages and make
the efficiency look too low. We have calculated the work of
acceleration and found it so low (about 25 mkg) that it can be
neglected.

It therefore seems justifiable to conclude from the results of
the above experiments that the oxygen requirement of 3 minutes
of work as a rule is greater when it is determined from the total
oxygen uptake of work and recovery than when it is calculated
from the oxygen uptake per minute in the steady state. The dif-
ference is caused by the fact, that the oxygen repaid in recovery
exceeds the oxygen deficit contracted at the start of the work.

A further consequence of this will be that the efficiency of a
short spell of work must be lower than the efficiency of a work

204

DRUNG ASMGSSEN.

lasting longer. This could be demonstrated in experiments in which
the work lasted from 0.5 min to 4 a 6 min. Two such series of
experiments were performed, one with tbe work intensity of 690
mkg/min and one with 1,150 mkg/min. The results of these ex-
periments .are presented in figs. 3 and 4.

It will be seen that witb the exception of the very short work
of 0.5 min duration the different values of the total oxygen
uptake of work and recovery lie on a straight line which, however,
does not pass, through the zero point. A mathematical expression
for the two lines in our experiments will be: total 02=1.30X4-0.15
for the work of 690 mkg/min and: total 02 = 2.17X -}-0.25 for
the work of 1,150 mkg/min where X represents the duration of
work.

The oxygen requirement per minute will consequently be:

Oo/min = 1.30

0.15

liters.

0.25

and: Oa/min = 2.17 liters.

The curves corresponding to these expressions are drawn in the
upper part of figs. 3 and 4, and around them are seen the actually
determined values of oxygen consumption per minute of work.

The curves as well as the formulas show, that as time (X)
increases the oxygen requirement per minute work will diminish,
approaching a value of 1.30 1/min and 2.17 l/min, respectively.
These values, of course, are the steady state values, as was also
confirmed by experiments. (Steady state values found: 1.30 1/min
and 2.16 1/min, respectively.)

The equations are not valid when the time is very short, as the
oxygen requirement of work lasting 0 minutes of course is 0.
For the very small values of X the oxygen stores of the organism
may play a role, making the work less anaerobic than later on.
This is maybe the explanation of the relatively low values of
total oxygen requirement for works of 0. 5 min duration in figs.
3 and 4.

From the 02 /min the net efficiency can be calculated. With an
assumed caloric value of oxygen of 4.8 it is found to vary in both
series from about 22 pCt at the shortest work up to about 25 pCt
for work of 5 min. duration.

206

EBIilKG ASMUSSEN.

Discussion.

The general conclusion that can be drawn from the experiments
in which the circulation to the legs was blocked, is that when the
energy production in rest or during work is forced to proceed via
anaerobic processes, mth a delayed aerobic recovery, the total
costs in oxygen are greater than when the energy is liberated
by simultaneous oxydative processes.

In the rest experiments the difference between oxygen deficit
during blockade and excess oxygen uptake during recovery is
first apparent when the circulation to the legs has been stopped
for more than 5 to 6 minutes. The deficit up to this time will be
about 100 cc of oxygen, and it seems justifiable to believe that
it is covered by the oxygen stores of haemoglobin and myoglobin
available in the legs. Uor periods of occlusion lasting more than
5 to 6 minutes true anaerobic conditions must prevail, and the
energy necessary to maintain life must be derived from anaerobic
processes. Of what kind these are cannot be decided, but the
slight increase in the blood lactates after about 20 minutes of
occlusion suggests that a glycogen breakdown has occurred.

IVhen the blocking of the circulation to the legs was performed
during woric the energy requirement is considerably greater, and
practically all of it must be derived from anaerobic break-downs,
presumably also of glycogen as the increased blood lactates in-
dicate. The oxydative recovery demands a comparatively large
amount of energy as judged by the high excess oxygen intake
during recovery. A production of mechanical energy that follows
the line; anaerobic break-do wn->-aerobic recovery, therefore, seems
far less efficient than one in which energy is derived by a more
direct oxidative process. The efficiency of the first process is —
judging from these experiments — only 40 to 70 pCt of the effi-
ciency of the latter.

This result may be compared with the results from myothermal
measurements showing that the “initial heat” and the “recovery
heat” are of the same order of magnitude, i. e. that the oxidative
resynthesis of substances broken down anaerobically has an
efficiency of only about 50 pCt (Meyerhof (1930)). The efficiency
of a muscle working anaerobically with a subsequent aerobic
recovery is consequently only about half as great as the efficiency
of anaerobic work. It therefore seems justifiable to conclude from

AEROBIC RECOVERY AFTER ANAEROBIOSIS.

207

OUT experiments that the anaerobic processes — at least the
“lactacide” ones — are shunted out from the recovery metabolism
of muscles working aerobically under normal conditions, and that
the oxidative energy is used more directly in rebuilding the po-
tential energy, thus enabling the muscle to work with the same,
high efficiency as in the anaerobic condition. The discrepancy
which existed between the absolute values found for the net
efficiency of isolated frog’s muscles and the efficiency determined
in.work experiments on man will hereby disappear: Under the most
favourable conditions the mechanical efficiency of frog’s muscles
working anaerobically was found to be 23 to 20 pCt, making the
efficiency of anaerobic work with subsequent aerobic recovery
12 to 10 pCt. This value is far below the value usually found in
experiments on man, but is of the same order of magnitude as
found in the present experiments, in which the muscles actually
worked anaerobically with a subsequent aerobic recovery.

Assuming the lag in oxygen uptake in the initial stages of work
to be the sign of a partly anaerobic period of energy production
one must expect to find an extra amount of oxygen taken up
during the time of recovery after work. A recovery from anaero-
biosis is possible during aerobic work, as we found in the experi-
ments in which the circulation to the legs had been- blocked, but
neither in our experiments nor in those Icnown from the literature
are there any signs of a “hump” on the oxygen curve in the first
minutes of the steady state. All available evidence seems to indi-
cate that the paying-back is postponed till after cessation of work.
A small part of the oxygen taken up after work will undoubtedly
be used in building up the oxygen stores of the organism, another
small part will be used by the heart and the respiratory muscles
in the transitory stage, but the rest no doubt is used in building
up the chemical stores of energy broken down anaerobically at
the onset of work. The' amount of oxygen taken up after the end of
work was in o'Ur experiments (with one exception) at all grades
of work larger than the oxygen deficit at the beginning of work
as might be expected from the point of view put forward above.
Corresponding results have been recorded a. o. by Herxheimer
(1935) Nielsen and Hansen ((1937) in a curve) and quite re-
cently by Bsiqldsen (1945). E. Hansen (1934), on the other hand,
claiim that the oxygen deficit at the start and the oxygen debt
repaid during recovery are exactly equal. Against his conclusions
may be objected that the grades of work he investigated were too

208

EBLING ASMUSSEN.

severe, so that a true steady state was not attained: Both oxygen
deficit and oxygen repayment increased with the duration of
work. Secondly, he determined the oxygen debts by planimetry
of curves drawn through points from several experiments. This
no doubt in itself may give rise to errors, as even the most careful
placing of the curve in relation to the individual points may
depend on an estimation, and further it must be borne in minH
that the oxygen uptake from the lungs at any given moment does
not correspond to the simultaneous oxygen consumption of the
muscles. Some factors will tend to make it too large (e. g. the
arterialization of depot blood) while others will make it too small
(e. g. the inevitable lag between oxygen usage in the muscles and
return of the venous blood to the lungs). The form of the curve
representing the pulmonary oxygen uptake, upon which depends
the areas measured by planimetry, thus becomes less well defined.
We are inclined therefore, to believe that the small differences
between oxygen deficit and oxygen repayment have been over-
looked in the experiments of E. Haiisen, and the more so as the con-
sequences of a larger oxygen repayment i. e. the lower efficiency
of work of short duration as compared with longer lasting work,
becomes evident from a scrutiny of E. Hansen’s data.

The suggestion that the efficiency of work is lower for short
spells of work was first put forward by Simonson and Hebesteeit
(1930), later advocated by Simonson and Sirkina (1934). The
increase in efficiency found was, however, very considerable, 200
to 600 pCt, and serious objections against their choice of work have
been raised by E. Hansen (1933) as well as by Crowden (1934).
Both E. Hansen and Crowden could show that the net efficiency
of a short spell of work is practically the same as for work of
longer duration, but as mentioned above, a scrutiny of B. Hansen’s
results discloses a difference in the efficiency of the same order
of magnitude as found in the present experiments although this
difference was not statistically significant. In Crowden’s experi-
ments — with one exception in which the trend was the same
as in our experiments — the technique of work (intermittent work
compared with continuous work) no doubt will make it utterly
hard to demonstrate small differences as those found in the
present experiments.

It seems justifiable, therefore, to conclude that the oxygen
repaid during recovery after light and moderate work is slightly
larger than the debt contracted at the start of work, thus making

AEROBIC RECOVERY AFTER ANAEROBIOSIS. 209

the efficiency of short spells of work lower than the efficiency
of work of longer duration.

It must be expected that in very short and severe work, in which
the anaerobic part of the work is more dominant the efficiency
will be considerably lower. An indication of this may be seen in
the upward slope of the curve representing the oxygen requirement
of running at increasing speeds presented by Hill (1926) although
in such cases also the increased work of stabilization and of res-
piration etc. will play an important role.

Summary.

The efficiency of the oxidative recovery after anaerobiosis has
been studied. The anaerobic conditions studied were those pre-
vaiUng in the legs when the circulation is cut off by means of
pneumatic cuffs round the thighs, and the partly anaerobic con-
ditions occurt'ng in the initial stages of work.

It was found that both in rest and in work the oxygen repay-
ment after the circulation to the occluded legs has been restored
is considerably in excess of the oxygen debt contracted during
the occlusion. The efficiency of work in which the energy is pro-
duced anaerobically with a subsequent aerobic recovery is there-
fore low; judging from these experiments only 40 — ^70 pCt of the
efficiency of aerobic work.

It is concluded from this that the anaerobic phases — at least
the formation of lactic acid — in the after-contraction metabol-
ism of muscle is shunted out during aerobic work, the oxydation
energy being utilized more directly in recharging the contractile
mechanism. In this way the loss in efficiency of about 50 pCt
which a delayed aerobic recovery involves is avoided. There is no
discrepancy between the values of efficiency found in experiments
on frog’s muscles and on human subjects.

In agreement with this it is found that the excess oxygen taken
up after cessation of a work in which a steady state could be reached,
is slightly greater than the oxygen deficit contracted at the
start of the work. A consequence of this is that the efficiency as
computed from the total oxygen uptake of work and recovery, is
lower than the efficiency calculated from the steady state values,
and further that the efficiency of a short spell of work is slightly
lower than the efficiency of work of longer duration.

14 — ^60215. Acta phys. Scandinav. Vol.lt,

210

EULING asmussex.

It must be expected that very severe work, wbicb is performed
mainly anaerobically with a subsequent aerobic recovery, bas a
low efficiency.

Beferenoes.

Asmussen, E., E. H, Christensen, and M. Nielsen, Skand. Arch.
Physiol. 1938. 79. 32.

— Ibidem 1939. 81. 201.

— Ibidem 1939. (b). 82. 212.

Asmussen, E., and 0. Boje, Acta physiol, scand. 1945. 10. 1.
Crowden, G. P., J. Physiol. 1934. 80. 394.

Edwards, H. T., J. biol. Chem. 1938! l25. 571. .

Eskildsen, P. P., Arbejdsfysiologiske Undersogelser (with an English
summary), Copenhagen 1945.

Hansen, Em. Arbeitsphysiol. 1933." 7. 801.

— Ibidem 1934. 8. 151.

Herxheimer, H., Ibidem 1935. 8. 801.

Hill, A. V., Muscular Activity, London and Baltimore 1926.
Meyerhof, 0., Die ohemische Vorgange im Muskel, Berlin 1930.
Nielsen, M., and 0. Hansen, Skand. Arch. Physiol. 1937; 76. 37.
Simonson, E. and Hebestreit, Pfliig. irch. ges. Physiol. 1930. 225.
498.

Simonson, E., and G. Sirkina, Arbeitsphysiol. 1935. 8. 560.

Biological InstiUito, Carlsborg Foundation, Copenhagen.

Interaction Between Fibrinogen and
Polysaccharide Polysnlfnric Acids.

TAGE ASTRDP nnd J0RGEN PIPER.
Received 21 December 1915.

During investigations on the properties of synthetic poly-
saccharide polysulfuric acid esters (Piper 1945 b), it was observed,
that some of the substances precipitated the fibrinogen in plasma.
As this was rather unexpected we have made an investigation on
the interaction between proteins and polysaccharide polysulfuric
acids. Some of the results obtained so far have been mentioned
briefly by Astrot nnd Piper (1945 b).

The substances investigated were prepared ns described pre-
viously (Astrup, Galsmar and Vodkert 1944, Astrup and Piper,
1945 a).

Exporiinental.

A. Fibrinogen in Plasnm and Cdluhse Snlfuric Acid.

The first problem was whether the precipitation of fibrinogen by
means of cellulose trisulfuric acid was quantitative. For analytical
investigations of proteins this would be of great significance, as cellulose
trisulfuric acid contains neither nitrogen nor phosphorous. A sample
(C-12) of a cellulose trisulfuric acid insoluble in concentrated salt solu-
tions was used. Citrate plasma was used (9 parts of blood -f- 1 part of
3.5 per cent sodium citrate),

^ Tdble_l shows two important things, i) No precipitate is obtained in
citrate plasma, when adding a sufficient surplus of C-12. 2) The resulting
solution forms no precipitate when heated to 56° for 10 min., from which
follows that the fibrinogen present is influenced by a surplus of cellulose
sulfuric acid in such a manner that it is not denaturated by heating as
it usually is. The absence of a precipitate after heating of the centrifuged
solutions (after precipitation with C-12) is therefore no indication of a

•212

TAGE ASTKirP AND J0IIQEN PIPER.

Table 1. (Exp. 3).

To Z ml ciirate plasma are added Z ml of solutions of G-IZ {as

sodium salt).

ResnUing concentration
of C-12 in per cent

After standing
for 15 min.

After centrifnging and
heating of the liquid
to 56’ in 10 min.

■water clear

almost clear

0.5

tnrbid

slightly tnrbid

0.2

turbid and precipitate

> >

0,1

bulky precipitate

> >

0.05

> >

O.ol

precipitate

> >

0.001

tnrbid

tnrbid and precipitate

O.oool

clear

> > >

quantitative removal of tlie fibrinogen 'with the precipitate, as previously
assumed (Piper 1915 b).

In Table 1 a concentration of about O.i per cent of C-12 gives the
best conditions for the precipitation, bur further experiments show
that the amount of fibrinogen precipitated, measured in terms of the
nitrogen content of the precipitate, varies from sample to sample even
under equal conditions. The influence of the medium on the precipita-
tion was therefore studied. It was thus found that dilution of the plasma
with physiol. NaCI, or still more with water, decreases the amount of
precipitate formed and eventually completely inhibits its formation.
The same was the case when increasing the salt content of the solution.
Table 2 and 3.

These experiments show that the precipitation is influenced to a
large extent by the composition of the mixture, and that this may be
the cause of the unsatisfactory results. Other substances were therefore
tried and compared with C-12.

Table 2. (Exp. 16).

To 1.2 ml citrate plasma are added 1.2 ml of HnO or solutions of NaCl
and then 0.24 ml of 1 per cent C-IZ. Standing for 10 min.

Solution

Result

HjO

bulky precipitate

0.5 per cent NaCl

> >

1.0 > 3 »

> >

2.0 » > >

clear

4.0 > > »

lO.o > » >

20.0 > » >

KIBKIXOGEX ;\X1) POTirSACCHARIDE VOhySVI.nmiC ACIDS. 213

Tnblo (Exp. J8).

To 1 ml cilraie plasma is added varying amounts of water and O.s
per cent solution of C-12. Standing for 10 min. First section with
a constant amount of C-7S, second section with a constant concentra-
tion of C-12.

JKO

O.fi per cent (,

-12 1 llesnlt

0 ml

0.2 ml

' linlky precipilnlc

1 >

0.3 >

1 > >

2 >

0.2 .>

j > > , ttirliid

3 ;

0.2 .

1 precipitate, trirljid

■1 .

0.2 >

’ > >

r, >

0.2 >

j tnrtiil

8 >

0.2 >

) cl Mr

0 ml

0.2 ml

i biilkyprccipitntc

O.-t >

1 > >

2 .

0-G i

j prccipit.atc

3 .

0.8 »

1 clear

5 .

1.2 »

j >

8 ^

1.8 .

1 *

B. The Precipitation of Fibrinogen in Plasma by Afeans of Other Sub-
stances.

The otbor sukstaiices tried were Liquoid-Roelic (Liq. R.), a starch
polysulfuric acid (S-3), a cellulose sulfuric acid soluble in concentrated
salt solutions (K-83.3) and a sulfuric acid ester of ccllulo.se glycollic
acid (K-81.1). Chitin sulfuric acid yielded no precipitate and was thorc-
foTO not investigated. Rabbit citrate plasma ^vns used throughout, this
investigation. The following symbols were used: 0 — clear, 1 = slightly
turbid, 2 = turbid, but no precipitate, 3 = minor precipitate, i = pre-
cipitate and 5 = bulky precipitate.

Table 4 shows that the ability to preeijutate fibrinogen in citrate
plasma is influenced to a different degree by the various substances used,
iho concentration of the substance is far more critical for Liquoid-
Roche than for any other of the substances investigated, Neither S-3,
i same amount of precipitate ns C-12.

dilution of the solution (keeping the concentration of the precipitatimr
substance constant) yields smaller amounts of precipitate (Table d)

f°«spntration inhibits the formation of a precipitate
Table G, but hero Liquoid-Rocho seems less affected. If howcver^ItrCl’
IS used instead of NaCl a precipitate is formed in all samples (also^the
diluted), but only when the MgCh-solution is added o/tertiie addition

TAGE ASTRUP AND J0RQEN PIPER.

Table i. (Exp. 22).

To one ml citrate plasma is added an amount of the substance dissolved
in water and the appearance is observed after standing for JO min.

lul

ntion

percentage

C-12

Liq.-R.

B:-83.2

K-81.1

1.0

2.0

• 1

0.1

■■

Table 5. {Exp. 24).

Plasma

Solution
ml, 1 per cent

HjO

C-12

Liq.-R .

8-3

K-83.2

K-8L1

1.0

O.l

1.0

0.2

1.0

0.3

2.0

1.0

0.5

4.0

1.0

0.9

8.0

By addition of NaOH to citrate plasma a precipitate iritli C-12 is
obtained even at pH 9.2, but at pH 9.7 the solution is almost clear.
It is very unexpected that a precipitate is obtained at such an alkaline
reaction. By addition of HCl to the plasma the amount of precipitate
after adding C-12 decreases and at pH 5.5 the mixture is only turbid.
At still higher degrees of acidity (pH 4. 2 ) a precipitate again appears,
this time due to a precipitation of plasma proteins by means of acid,
probably in connection with the amount of C-12 present (see later). In
oxalate plasma the precipitate appears as finer particles than in citrate
plasma.

The precipitate formed by adding C-12 to plasma is almost completely
redissolved by addition of a surplus of the substance. An exception is
K-81.1, where no resolution seems to occur. This is also the case when
diluting a precipitated plasma mixture with water. But when adding a
surplus of sodium chloride solution, it is the precipitate formed by
addition of Liquoid-Eoche which does not redissolve. An addition of
heparin in not too large amounts to plasma before adding C-12 does not
inhibit the precipitation.

FIBRINOGEN ANB POLYSACOHARIBE POLYSULFURIC ACIDS. 215

Table 6. (Exp. 26).

To one ml plasma is added 0,5 ml of different concentrations of
NaCl and 0,1 ml 1 per cent solution of the substance.

NaCI-soIntion

Percentage

C-12

Liq.-R.

S-3

K-83.2

K-81.1

O.O

1.0

2.0

5.0

“4

lO.o

C. The Stabilisation of Fibrinogen in Plasma by Means of Polysaccharide
Sulfuric Acids.

As pointed out in section A the presence of 0-12 in citrate plasma
inhibits the denaturation and precipitation by heating to 56° of the
fibrinogen not precipitated previously by the addition of C-12 to the
sample. This was unexpected because fibrinogen is the most labile of
the plasma proteins and one of the proteins which is most easily dena-
tured. This phenomenon was therefore investigated more elosely.

First the action of heparin was tried, and it was found that increasing
amounts of heparin decreased the amount of precipitate formed after
heating the solution to 56° for 10 min. The effect is, however, inferior to
the effect of C-12, as a mixture of 1 ml citrate plasma and 1 ml 5 per cent
heparin (»Leoi>) still becomes turbid after heating, while addition of 1 ml
2 per cent 0-12 yields almost clear solutions. The centrifugate from
plasma (2 ml) with C-12 (0,2 ml 1 per cent solution) added is clear after
heating to 56° (10 min.), but a slight precipitate appears after addition
of saturated ammonium sulfate to 0.3 saturation of such a centrifugate
(before heating) and also after dilution with water and acidulation with
acetic acid. Such processes are therefore not inhibited by C-12, but
neither are they specific for fibrinogen, blood serum showing the same
properties. CHtin disulfuric acid (K-51) acts in a manner similar to
that of heparin. Two substances, with neither fibrinogen precipitating
properties nor anticoagulant activity, viz. a cellulose glycollic acid
(E-96) prepared as described previously (Astrup and Piper 1945 a)
and a chitosan glycollic acid (K-152.12) to be described in a subsequent
paper (Astrup, Bars0e and Piper) were tried. None of these substances
showed any definite inhibitory effects on the heat denaturation.

D. The Precipitation of Purified Fibrinogen Solutions.

In order to exclude interference with the precipitation of fibrinogen
rom other plasma proteins, purified fibrinogen solutions prepared
according to Astrup and Darling (1942) from ox plasma were next
investigated. The fibrinogen solution was used diluted with physiol
NaCl in the proportion 1:3.

216

TAQE ASTBUP AND J0RGEN PIPER.

To 10 ml fibrinogen solution is added O.i ml 1 per cent C-12. A bulky
precipitate is formed and removed by centrifugation. Addition of am-
monium sulfate (0.3 saturation) to the solution yields no precipitate,
and the same is the case after heating to 56° for 10 min. Addition of
NaCl (1 ml 10 per cent NaCl to 2 ml fibrinogen solution and 0.2 ml I per
cent C-12) inhibits the formation of a precipitate, and a precipitate
formed previously is redissolved. The addition of serum instead of
physiol. NaCl inhibits to a certain degree the formation of a precipitate:
1 ml fibrinogen solution (diluted) 3 ml physiol. NaCl 4-0.4 mil per
cent C-12 -* precipitate. 3 ml blood serum instead of 3 ml physiol.
NaCl -»■ only slight opalescence. But 3 ml plasma yields a precipitate;
in this case, however, the amount of fibrinogen present is also increased.

A chitin disulfuric acid (K-51) precipitates fibrinogen from a fibrin-
ogen solution, contrary to what is the case with fibrinogen in plasma
(1 ml fibrinogen -f- 3 ml physiol. NaCl 4- 0.4 ml 1 per cent K-51). The
same is the case to some extent with a chondroitin polysulfuiic acid
(K-7lr), which in fibrinogen solutions produces turbidity. Liquoid-Eoche
precipitates a fibrinogen solution. A surplus of cellulose trisulfuric acid
(C-12) precipitates a fibrinogen solution, and the precipitate is not
dissolved again when adding an excess of the substance (contrary to the
conditions in plasma). Large amounts of heparin inhibit the formation
of a precipitate (1 ml fibrinogen solution -j- 3 ml 0. 7 5 per cent heparin
no precipitate or turbidity. 1 ml fibrinogen -f 3 ml HjO -f- 0.4 ml 1 per
cent C-12 -* precipitate. 1 ml fibrinogen 4- 3 ml 0.7 5 per cent heparin
4-0.4 ml per cent C-12 -♦no precipitate). The sulfuric acid ester of
cellulose glycollic acid (Astrup and Piper 1945 a) precipitates fibrino-
gen in plasma, but a similar substance prepared by Kareer, Koenig
and Usteri (1943), and kindly furnished us by Professor Dr. Paul
Karrer, Zurich, did not yield any precipitate in plasma, cf. Piper
(1945 b). In fibrinogen solutions, however, a precipitate appears. The
cellulose glycollic acid itself produces no precipitate. At slightly alkaline
reaction (phosphate buffer) no precipitation occurs when adding cellu-
lose trisulfuric acid or chitin disulfuric acid, the pH value depending on
the concentration of buffer salts present.

Thus purified fibrinogen solutions are more easily precipitated than
fibrinogen in plasma. Heparin, which as mentioned inhibits the forma-
tion of a precipitate when cellulose suKuric acids are added, also acts
stabilizing on the heat denaturation of fibrinogen (1 ml cone, fibrinogen
solution 4-3 ml 0.75 per cent heparin -> no precipitate after heating
to 56° for 10 min.). Cellulose glycollic acid, which also yields no precip-
itate, but has no anticoagulant property, inhibits the denaturation
only to a small extent. Fibrinogen itself in phosphate buffer at slightly
alkaline reaction (pH ~ 8) is not precipitated by heating to 56° for 10
min.

E. Action on Other Proteins.

While an interaction between proteins (fibrinogen) and polysacchar-
ide sulfuric acids at neutral or slightly alkaline reaction was unex-
pected, it was to be assumed that these high molecular acids would

FIBRINOGEN AND POIiTSACCHARIDE POLYSULFURIC ACIDS. 217

react with proteins in general at slightly acid reaction in the same
manner as nucleic acids and heparin. This proved also to be the case.

Ovalbumin (recrystallized) (1 ml) in O.i-n acetate buffer (5 ml) begins
to form precipitate at pH 4.4 when adding 1 ml 1 per cent cellulose tri-
sulfuric acid (C-12) and at pH 4.1 when adding chitin disulfuric acid
(K-51). A precipitated pig globulin under the same conditions begins
to precipitate at pH 5.3 with C-12 and at pH 5.0 with K-51.

While the polysaccharide sulfuric acids protected fibrinogen against
denaturation at 56°, rabbit serum (1 ml), with 3 ml IVs per cent heparin,
C-12 or K-51 added, showed only insignificant differences when heated
to 90° for 10 min. The inhibitory effect on the heat denaturation of
other proteins is therefore not very pronounced.

Discussion.

The experiments described indicate that the interaction be-
tween fibrinogen and polysaccharide sulfuric acids is of a rather
specific nature. In general, high molecular acids only react with
proteins on the acid side of the isoelectric points, this is the case
of nucleic acids, E. Hammarsten (1924), and heparin, Fischer
(1935), Jaques (1943). While it is to be assumed, that other
polysaccharide sulfuric acids than heparin act in a similar manner
on proteins, it was unexpected that a reaction with fibrinogen also
occurs on the alkaline side of the isoelectric point. This interaction
was found to be highly dependant on the composition of the react-
ing medium, thus indicating a more loose combination between
fibrinogen and the high molecular acids, than is the case in the
compounds resulting from proteins and high molecular acids at
an acid reaction, and which are salt-like products. The precipitate
in question probably is a coacervate.

How, fibrinogen is a protein of unusual properties originating
from its structure as a thread-shaped protein of high molecular
weight. Holmberg (1944) describes fibrinogen as a thread molecule
with a molecular weight of about 700,000. The axial ratio of the
molecule is 60. According to Cohn and coworkers (1944) it has a
molecular weight of about 500,000 with a length of 900 Angstrom
and a diameter of 33 Angstrom (axial ratio = 30). The poly-
saccharide polysulfuric acids likewise are high molecular electro-
lytes. Presumably even at slightly alkaline reaction some positively
charged groups on the fibrinogen molecule (arginine and lysine
residues) may be able to combine with the negative charged
polysaccharide sulfuric acid to such an extent that a precipitation

218

TiQE ASTKUP AND J0RGEN PIPER.

takes place, although the acidic groups (dicarboxylic acid residues)
are negatively charged at this pH value. The presence of such
charged groups in the compound formed may explain the de-
pendence of the precipitation on the composition of the mixture
as disclosed in the present investigations. The thread-shape of
the molecules of fibrinogen and of the cellulose derivatives may
facilitate the formation of the compound, while more globular
proteins probably do not react to the same extent. Polysaccharide
polysulfuric acids (or similar compounds of high molecular weight)
of more globular configuration possibly may not precipitate fibrin-
ogen in plasma. According to Gronwall, Ingelman and Mosi-
MANN (1945) heparin has a molecular weight of only 17,000, but
shows a high sfrictional ratio)) (2,5) indicating an asymmetric
shape.

In a previous paper, Piper (1945 a), it was found that the syn-
thetic polysaccharide polysulfuric acids agglutinate the blood
platelets, and this phenomenon was later associated with the
interaction between the polysaccharide derivatives and fibrinogen.
Piper (1945 b), but as it also was found that a substance (»Liquoid-
Roche») was able to precipitate fibrinogen without agglutinating
the blood platelets, this is not the whole explanation, even if the
precipitation alone is sufficient to characterize such a substance
as toxic. An excess of C-12 or K-51 still agglutinates the platelets,
even if no precipitate is formed. A suspension of kaolin or bacteria
is only agglutinated by C-12 in the presence of fibrinogen. Also the
complement-inactivating properties of such substances may be
taken into consideration, Wilander (1939).

The fibrinogen precipitating properties explain why Astrup, Galsmar
and VoLKERT (1944) obtained straight parallel lines in investigations
on the clotting of recalcified oxalated plasma containing various
amounts of cellulose trisulfuric acid, while heparin, and to some extent
also chitin disulfuric acid, yielded curved lines. Similar results were at
the same time obtained by Bertrand and Qutvy (1944) using oLiqnoid-
Eochcj). Curves of such appearance may be obtained when clotting
plasmas containing various concentrations of fibrinogen (BarS0E and
Sels0).

While the fibrinogen precipitating properties of polysaccharide poly-
sulfuric acids have not been known before, it was found by Stuber and
Lang (1932) that »Liquoid-Roche» precipitates fibrinogen. The same
was found by ZuNZ, Mena-Ugalde and Vesselovsky (1934, 1935),
who, observed a fall in the number of platelets. According to Chorine
(1941) fibrinogen is not the only protein of plasma, which is precipitated
by means of Liquiod. Calcium ions are of importance for this reaction.

FIBRINOGEN AND POLYSACCHARIDE POLYSULFURIC ACIDS. 219

According to Mylon, Winternitz and de Suto-Nagy (1942) also
protamin selectively precipitates fibrinogen from plasma, which is not
quite consistent with our results concerning the action of clupein sulfate
on fibrinogen solutions (Astrup 1944). According to Chargaff and
Bendich (1943) ninhydrin and some naphthoquinone sulfonic acids
produce a coagulation of fibrinogen resembling the formation of a natural
fibrin clot.

Fischer (1935) found that heparin, eontrary to nucleic acids, protects
proteins against denaturation by means of acids, and that even the
heat denaturation may be hampered, although, according to Colldahl
and Kahlson (1939), not to the same extent as by using other anticoag-
ulants (chlorazol fast pink). Our investigations showed that the syn-
thetic polysaccharide sulfuric acids onl}’' to a minor degree inhibited
the heat denaturation of proteins.

These investigations were supported by tlic »P. Carl Petersens
Fond)), and the rabbits used were placed at our disposal by »Lovens
ketJiisl-e Fabrih), Copenhagen.

Summary.

The interaction between fibrinogen and polysaccharide poly-
sulfuric acids is studied, and it is found that the formation of a
precipitate is influenced to a large extent by the composition of
the mixture (concentration, salt content, pH) and the nature of
the substances used. Even on the basic side of the isoelectric
point a precipitate is formed, which may be due to the properties
of fibrinogen and the polysaccharide derivatives as thread-shaped
high molecular substances. The precipitate formed probably is a
coacervate. These observations explain the toxicity of the sub-
stances in question.

Literature.

Astrup, T., Acta Physiol. Scand. 1944. 7. Supjil. 21 . 108.

Astrup, T., and S. D.^bling, Ibidem. 1942. 4. 45.

Astrup, T., I. Galsmar and M. Volkert, Ibidem. 1944. S. 215.
Astrup, T., and J. Piper, Ibidem. 1945 a. 9. 351.

Astrup, T., and J. Piper, Nord, Med. 1945 b. 2S. 2405.

Barsoe, 0. C., and S. Selso, (to be published).

Bertrand, L, and D. Quivy, R. C. Soc. Biol. Paris. 1944. 138. 404.
Chargaff, E., and A. Bendich, J. Biol. Chem. 1943. 149. 93 (cited from
An. Rev. Physiol. 1944. 6. 298).

Chorine, V., C. R. Soc. Biol. Paris. 1941. 135. 451. 454.

220

TAGE ASTRUP AND J0RGEN PIPER.

Cohn, E. J., J. L. Oncley, L. E. Strong, W. L. Hughes and S. H.

Armstrong, J. Clin. Invest. 1944. 2S. 417.

CoLLDAHL, H., and G. Kahlson, Skand. Arch. Physiol. 1938. 78. 149.
Fischer, A., Biochem. Z. 1935. 27S. 133.

Gronwall, a., B. Ingelman and. H. Mosimann, Upsala Lakareforen.

fork. Ny F. 1945. 50. 397.

Hammarsten, E., Biochem. Z. 1924. 144. 383.

Holmberg, C. G., Ark. Kemi, Mineral. Geol. 1944 . 17 A. Nr 28.
Jaques, L. B., Biochem. J. 1943. 37. 189.

Karrer, P., H. Koenig and E. Usteri, Helv. Chim. Acta. 1943. 26.
1296.

Mylon, E., M. C. Winternitz and G. J. de Suto-Nagy, J. Biol. Chem.
1942. 143. 21.

Piper, J., Acta Physiol. Scand. 1945 a. 9. 28.

Piper, J., Farmakologiske Undersogelser over syntetiske koagulations-
hsemmende Stoffer af Heparingruppen. Copenhagen. 1945 b.
Stuber, B., and K. Lang, Biochem. Z. 1932. 244. 214.

WiLANDER, 0., Skand. Arch. Physiol. 1939. 81. Suppl. 15.

ZuNZ, E., C. Mena-Ugalde and 0. Vesselovsky, C. E. Soc. Biol.
Paris. 1934. 116. 336.

ZuNz, E., C. Mena-Ugalde and 0. Vesselovsky, Le Sang. 1935. 9. 124.

From the Department of Physiology, University of Lund.

The Influence of g-Stropliantin on tlie Mech.an-
ical Properties of Cardiac Muscle.

GUNNAR liUNDIN.

Received 24 December 1945.

It is generally accepted that the glucosides of digitalis improve
the activity of a failing heart. How this improvement is effected
is still an unsettled question.

According to, among' others, Mackenzie (1918) and Leivis
(1937) the efficacy of digitalis is attributed to a depression of the
auriculoventricular conduction. Christian (1933), Wencicebaoh
(1930) and others attribute the effect to an increase of the con-
tractile power of the failing heart.

In the investigations of the latest years we find the same dua-
lism. Gold and Cattell (1940) describe a clear positive inotrop
effect of digitalis. Katz, Bodbar'd, Friend and Bottersi.ian
(1938) and Mo Michael and Sharpey-Schater (1944:) do not
find, however, any direct action of digitalis on the heart muscle.
The last-mentioned authors attribute the beneficial therapeutic
effect of digitalis to an action on the periferal vessels, chiefly those
of the liver. The divergent opinions derive from numerous ex-
perimental observations, mostly on hearts in situ of isolated
hearts.

In experiments on such preparations there will come in several
factors difficult to control, which will cause trouble in inter-
preting the results obtained. To be able to make a detailed
analysis of the effeqt of digitalis on cardiac muscle, one must
have a preparation as uncomplicated as possible. In these in-
vestigations we have used parallel threaded muscle bundles

222

QUNNAE IiUNDIN.

(0.2 — 0.4 mm tliick and 1.5 — 3 mm long) from the cardiac ven-
tricle of a frog. Such a preparation in an oxygen saturated Einger
solution can be considered to work under perfect aerobic con-
ditions. (Clark, Eggleton, Eggleton, Gaddie and Stewart
1938.) The purpose of these investigations is to analyse the effect
of g-strophantin on the mechanical properties of mormaU cardiac
muscle, that is strength of contractio7i, diastolic tension, elasticity,
and viscosity.

Method.

The small muscle bundles are taken both from the apex and the
base of the ventricle.

The muscle is prepared in ice-cooled Einger solution under a bin-
ocular microscope (enlarging about 20 times).

The composition of the Einger solution is as follows; 0.67 g NaCl,
0.02 g KOI, 0.04 g CaClj-f- 6 HjO, and O .02 g glucose in 100 cc aq .dest.
To ensure a suitable colloidosmotic pressure 3 g dextran is added.
(Gronwall and Ingelmann 1944). The Einger solution is buffered
with NaHCOj, and then a stream of a mixture of 1 % CO 2 and 99 %
O 2 is passed through, to a pH 7.2 — ^7.4. The temperature of the solu-
tion 6° — 1° is controlled thermo-electrically and remains constant
throughout the experiment. The duration of contractions will thereby
be sufficiently protracted to facilitate the determinations of the mechan-
ical properties.

The muscle bundle is stimulated at constant intervals, 5 per minute,
by means of a motor-driven switch in the outlet circuit of a multi-
vibrator stimulation device which gives rectangular current pulses of
different durations. In this case, as a rule, we have used impulses with
a duration of 5 milliseconds. The silver tweezers holding the ends of
the muscle bundle are used as stimulating electrodes and are electric-
ally isolated from the recording apparatus.

A modification of the condensor-myograph deviced by Buchthal
(1942) (fig. 1) has been used to record the variations of tension in the
muscle. A movable condenser plate [(2) fig. 1] approaches a fixed con-
denser plate at increased muscle tension during contraction or at
stretching. The capacity-variations are recorded by means of a high-
frequency apparatus with amplifier and electrostatic oscillograph
(fig. 2). The apparatus is also used for the determination of the dyn-
amic elasticity in vibration experiments.

The investigations of the above mentioned properties are made at .
lengths varying between 100 % and 180 % of equilibrium length. The
equilibrium length is the length at which the muscle, when stretched, just
begins to develop tension. At higher elongations the muscle must con-
solidate for more than one hour after the stretch, before the examination
of the mechanical properties can be performed. (Ltjndin 1944.) The

224

GUNNAR LRNDIN.

Pig. 2. Block diagram of apparatus for recording tension and stiffness.

deviation of the movable condenser plate caused by the stretching of
the muscle is compensated by leading a current tluough the solenoid
[(8) fig. 1] which moves in the field of a permanent magnet. The
compensation current is read off and 4 — 5 contractions are recorded.
The vibrations of the movable condenser plate are brought about by
sudden current impulses in the solenoid (8), both at the height of con-
traction and at rest. The springs (4, 5) which carry the movable system,
have such a stiffness, that during contraction the shortening of the
muscle does not amount to more than 0.5 % of its total length. This
length alteration is slight and the contraction can be regarded as iso-
metric. Extra tension during the contraction is measured directly on
the curves in millimeters, and an absolute measure is obtained by com-
paring the height of the contractions to the deviations from the base-
line which are produced by current impulses corresponding to well-
known loadings. The tension of the muscle at rest is given by the size
of the compensation current. The total tension during the contraction
is then the sum of the tension of the muscle at rest and the extra ten-
sion produced during the contraction.

In a number of esperiments the muscle is slightly released from
maximal tension during every other contraction. The decrease in length
of the bundle which, in each experiment, is constant ca. 10 — ^15 %
of equilibrium length, is brought about by current variations in the
electro-magnet [(15) fig. 1]. A camera, in which photographic paper
is transported ■ at constant speed, is used for the recording.

G-strophantin which has been chosen on account of its rapid effect,
is used in concentrations varying between 1 part in 5 • 10® and 1 part
in 2 • 10®. In the majority of the experiments the concentration 1 part
in 10® is used.

At the concentration 1 part in 5 • 10® — ^1 part in 10® several records
are made, beginning 5 minutes after the addition of g-strophantin and
finishing within 30 minutes. At higher concentrations records are
taken for a period of 20 minutes.

INFLUENCE OF G-STROPHANTIN.

225

Eesnlts.

Muscle tension at rest and during contraction.

Tension at rest is, with the exception of a few experiments with
a great initial length, the same before as after the addition of
g-strophantin. In a couple of experiments with an initial length
of 180 % of equilibrium length a fall in tension after g-strophantin
can be observed, but this fall is not greater than it can be ex-
plained as result of consolidation. In none of the experiments has
a fall in tension appeared which can be interpreted as a diastolic
effect of g-strophantin.

Extra tension during contraction does not increase after g-stro-
phantin but remains unaltered. This is the case at all elongations
investigated. The curve for isometric maxima coincides for drug-
ged and normal muscles. This is also the case at lower g-strophantin
concentrations at higher temperatures of the Ringer solution.
As a criterion of the effect of g-strophantin we only have the
toxic effect for the experiments in question. This, effect has been
observed in experiments with high g-strophantin concentrations

I part in 6 • 10® and 1 part in 2 • 10®, where it has appeared after
15 — 20 minutes as a fall in the strength of contraction and an
increase of the tension at rest. The changes in mechanical pro-
perties in this stage of the effect of g-strophantin are the objects
•of further investigations.

Stiffness at rest and during contraction.

Stiffness (S), which is the ratio between corresponding increase
A tension

in tension and length — and is measured in dyne cm

is an expression of the elastic properties. In a body with such
great variations in cross-section and length this quantity is more
suitable than the elasticity modulus.

In our experiments the dynamic stiffness is determined by
measuring the frequency of the recording system -{■ muscle. The
frequency of vibrations is adjusted so that it lies between 8 and

II cycles per second.

When vibration time and mass of the system are known, the
.stiffness of the muscle (Si) can be calculated from the formula

7t •

■So

■ 15 — iG0215. Acta phys. Scandinav. Vol. 11,

226

GUNNAR LUXDIN.

U M Aik i4 .iis liLjji.j4kjlL-A4_,
?. ifTTTtTTftriTTTT'fffST!?^ TflTfTTffl^TTTT*

• ■..r i.r c, . .. A'l'iV.v,-, -)■■', ',, t,VmhmTT7'|!!

Fig. 3. Two contractions, the upper befoie g-strophan tin, the lower 10 min.
after addition of g-strophantin 1 : 10'. Time maris 1/50 sec.

wKete So ~ stiffness of the system -without muscle, Ti vibration
time for the system + muscle and mo the mass of the vibrating
system. The mass of the muscle can be ignored.

At least four contractions -with" vibrations are recorded each
of them preceded by two vibration impulses at rest. The stiffness
thus found at every registration is the mean of eight measure-
ments. Stiffness is measured in 15 experiments both at rest and
during contraction, with and without g-strophantin.

Stiffness at rest does not show any change after g-strophantin
when referred to the same length, except in experiments at high
initial lengths, where the consolidation is not yet finished. The
fall in stiffness does not exceed that due to the fall in tension.

INFLUENCE OP G-STROPHANTIN.

227

Hence stiffness during rest in proportion even to tension is not
influenced by g-stropbantin.

Stiffness during contraction is the same before and after addi-
tion of g-strophantin, both when referred to the same tension and
the same length.

If we call the arithmetical mean of stiffness at rest before
addition of g-strophantin 100, we get the following values for
stiffness referred to the same length (15 experiments).

before g-strophantin after g-strophantin

rest contraction rest contraction

100 ±8 % 410 ±3 % 97.7 ±7 % 395 4:3 %

As stiffness at rest and during contraction is the same ivhen
referred to the same tension and as g-strophantin does not alter
the stiffness, we can conclude that also after g-strophantin stiff-
ness is identical when referred to the same tension.

Viscous properties.

Decrement of amplitude in vibration in the stiffness experi-
ments described above gives a measure of viscosity. If we call
the proportion between two successive vibrations (f), the so-called
damping constant (P) can be worked out from the formula

P = log D • ^

where T = vibration time in sec. and m = mass of the vibrating
system in gram. This is an expression for total damping. To
find the damping of the muscle (Pf), the damping of the regis-
tering system itself (P,,) must be subtracted from the total damp-
ing (P).

We get P — P^ = Pj

Viscosity in cardiac muscle, expressed by the damping constant,
when referred to the same length, is found to be the same before
as after the addition of g-strophantin. As was the case in the
stiffness and tension measurements in a couple of experiments
at high elon ations, we get a slight fall of viscosity at rest. This
decreasing viscosity agrees with the picture we have got of the

228

GUNNAK ITODIN.

changes in the mechanical properties by consolidation after an
elongation.

The arithmetical mean of viscosity at rest before addition of
g-strophantin being 100, we get the following values for viscos-
ity referred to the same length, measured in 12 experiments.

before g-strophantin after g-strophantin

rest contraction rest contraction

100 ± 12 % 382 ± 6 % 97 ± 12 % 362 ± 5 %

From the above experiments it is also obvious that viscosity
referred to the same tension does not undergo any change after
addition of g-strophantin.

If we release a muscle at rest or during contraction we get
a deep fall in tension. The tension attained dturing release-con-
traction lies considerably below the isometric tension at the same
length. This is due to the slow consolidation in cardiac muscle,
and is an expression of its viscosity. The fall in tension of the
muscle, referred to the same length and the same tension, is the
same before and after the addition of g-strophantin. Thus the
experiments with release-contractions show no changes in the
viscous properties of the cardiac mxiscle produced by g-strophantin.

It appears from the above experiments, that on the preparation
used, g-strophantin does not influence the diastolic tension or
extra tension during isometric contraction, at a stage in the ex-
periment when g-strophantin should have exerted its thera-
peutic action. Nor is stiffness at rest and during isometric con-
traction influenced by g-strophantin.

Discussion.

It might be possible that other results could be expected in
isotonic contractions due to changes in viscosity. A decrease in
viscosity could increase the effect of the isotonic contraction,
without any appreciable influence on the tension in isometric
contraction.

However, viscosity, expressed by the damping constant, does
not show any changes after addition of g-strophantin. This is
further supported by the release experiments. The fall in tension
during release gives an idea of the work in an isotonic contraction
before and after g-strophantin. The decrease in tension is then

INFLUENCE OF 6-6XR0PHANTIN.

229

inversely proportional to the work developed. The fall in tension
being the same before and after g-strophantin, Ave may conclude
that g-strophantin does not influence the capacity of work.

Thus these investigations show that g-strophantin does not
affect the mechanical properties of the normal cardiac muscle.
Before we can draw any conclusions with regard to the thera-
peutic value of the drug, the experiments have to be completed
with an investigation of hypodynamic cardiac muscle.

Summary.

The mechanical properties of cardiac muscle have been examined
on isolated parallel threaded muscle bundles under aerobic con-
ditions, before and after addition of g-strophantin. G-strophantin
does not influence any of the following properties: Diastolic ten-
sion, extra tension during contraction, stiffness at rest and during
contraction, viscosity or the capacity of work during isotonic
contraction.

References.

Buchthal, F., Dot Kgl. Danske Vidensk. Selskap. Medd. 1942. A7F. 2.
Christian, H. A., J. Amcr. med. Ass. 1933. 100 . 789.

Clark, A. J., M. G. Egoleton, P. Eggleton, E. Gaddie and C. P.

Steavart, The Metabolism of the Frog’s Heart. Edinburgh 1938.
Gold, H. and M. Cattell, Arch, intern. Med. 1940. 05 . 263.
Gronwall, a. and B. Ingelman, Acta Physiol. Scand. 1944, 7 . 97.
ELatz, L. M., S. Rodbard, M. Friend and W. Rottersman, J. P'liarma-
col. 1938. 62 . 1

Lewis, T., Diseases of the Heart, ed. 2. London 1937.

Lundin, G., Acta Physiol. Scand. 1944. 7 . Suppl. XX.

Mackenzie, J., Princijjles of Diagnosis and Treatment in Heart Af-
fections, ed. 3. London 1937.

McMichael, j. and E. P. Sharpey-Schafer, Quart. J. Med. 1944.
37 . 123.

Wenckebach, K. F., Brit. Med. J. 1930. 1 . 181.

From the Pharmacological and Physiological Departments, Karolinska
Institutet, Stockholm.

Efferent Impulses in the Splanchnic Nerve.

B. GERNANDT, G. LILJESTRAND and Y. ZOTTERMAN.
Received 2 Jannary 1946.

The paramount importance of the splanchnic area for the re-
gulation of the arterial blood pressure 'was demonstrated by
Ludwig and his school (Ludwig and Thiry, 1864, Cyon and
Ludwig 1866) and has been repeatedly confirmed, e. g., in modern
times by Jansen, Tams and Achelis (1924) and Kramer and
Wright (1932). Cyon and Ludwig also found that the effect
on the blood pressure of electrical stimulation of the central end
of the depressor nerve in the rabbit was greatly reduced after
section of both splanchnic nerves or after compression of the aorta.
In the dog, extirpation or severance of both ganglionic chains
has been found to abolish or greatly reduce the pressor reactions
after the clamping of the carotids (Bacq, Brouha and Heymans
1934, Schneider 1934), and the analysis carried out by Bern-
THAL, Motley, Schwind and Weeks (1945) has shown that the
chemoreflexes as well as the pressoreflexes elicited from the
sinus region have as their sole efferent pathway the thoracico-
lumbar autonomies. In the cat, however, pressoreflexes may
still be evoked, though to a reduced degree, from the sinus re-
gion, even after both the sympathetic ganglion chains have been
extirpated (Bacq, Bremer, Brouha and Heymans 1939). That
the reflex action from the buffer nerves on the blood pressure is
to a great extent exercised by altering the state of constriction
of the vessels within the splanchnic area has also been observed
directly. Thus electrical stimulation of the central end of the
depressor nerve has been seen to cause a dilatation of the vessels

EFFEREST IMPEI^ES IN TtlE SPEAXCUXIG NERVE.

231

of the kidney and the intestine (for reference of Heymans,
Bouckaert and Regniers 1933), and similar results have been
obtained after electrical stimulation of the sinus nerve in respect
of the vessels of the spleen, kidney, intestines and liver. Chemo-
reflex reactions from the carotid body have been found in the
spleen (Heymans, Bouckaert, Euler and Dautrebande 1932)
and in the intestines (Bernthal and Schwind 1945). Corres-
ponding effects have also been observed concerning the secretion
of adrenaline (for literature cp Euler and Liljestrand 1934).

The buffer nerves are stimulated by the intracarotid and intra-
aortic pressure, as well as by the chemical composition of the
arterial blood, both of which, imder physiological conditions,
reflexly influence the circulation (cp Euler and Liljestrand
1942) and probably the adrenaline secretion. There is a great
difference, however, between the modes of action of these two
kinds of stimuli. An increase in the intracarotid or intraaortic
pressure leads to a rise in the number of impulses in the buffer
nerves, as demonstrated by the action potentials, which in its
turn is followed by a lowering of the heart rate, a vasodilation and
a lessened secretion from the adrenal medulla. Oxygen want or
carbon dioxide accumulation also elicit an increase in the number
of impulses in the buffer nerves (cp Euler, Liljestrand and
Zotterman 1939), usually of much smaller amplitude, but the
ultimate results is quite the reverse; increased heart rate, rise
of blood pressure and probably increased secretion of adrenaline.
The interpretation generally accepted is that the impulses evoked
by stimulation of the pressor receptors cause an inhibition of the
sympathetic centers responsible for vasoconstriction, accelera-
tion of the heart rate and adrenaline secretion, whereas stimu-
lation of the chemoreceptors calls forth increased activity of
those centers. According to several authors (cp Heymans, Bouc-
kaert and Regniers p. 39), stimulation of the central end of the
depressor nerve not only causes a diminished tone of the vaso-
constrictors but also an increase in the tone of the vasodilators.
There will thus exist a kind of reciprocal innervation. A similar
arrangement might be expected for the sinus mechanism, though
no direct evidence seems to be available.

Since both presso- and chemoreceptors are constantly stimu-
lated under physiological conditions, the resulting tone of the
centers will be largely dependent on the relative magnitudes of
their influences.

232

B. GERNANDT, G. MUESTRASI) AXD T. ZOTTERMAX.

Technique and Procedure.

All our experiments have been performed on cats in chloralose
anesthesia, O.os g per kg body-weight being injected intravenously.
The action potentials from the efferent fibres of the splanchnic nerve
were recorded by means of an amplifier and the cathode-ray oscillo-
graph previously described (Zotterman 1936).

The Preparation. After removing the skin just below and above the
last rib, close to the vertebral column, the hluscul. latiss. dorsi and the
oblique abdominal muscles were transected and drawn apart. By cut-
ting through the inferior serratus posterior muscle parallel and close
to the last rib, the splanchnic nerves were exposed where they pass
over the diaphragmatic root. The nerve was laid free as far as its
entrance into the coeliac ganglion, where it was transected. Care had to
be taken at this moment that the nerve was not unduly stretched, as
the cats exhibited very violent reaction to the nerve section. In order
to reduce the signal-to-noise ratio, the common sheath of the greater
splanchnic nerve was pulled off from the cut end. This procedure can-
not, however, be applied to the lesser splanchnic nerve, as this nerve
generally splits up into several thin fascicles some distance before
entering the ganglion. In some preparations of the greater splanchnic
nerve there was a very high spontaneous electric activity, obviously
due to injury potentials set up by the afferent fibers. The behaviour
of these fibres has been subjected to special study, which is reported
separately (Gernandt and Zotterman 1946).

The efferent fibers in the splanchnic nerve of the cat seem to be
below 3 II in diameter. Action potentials from single efferent fibers
can thus be recorded only from very thin preparations which exhibit
a high signal-to-noise ratio (cp Zotterman 1936). The electric response
of the efferent fibers in our preparations consisted of summed up
potentials, and thus the records did not permit of any direct counting
of the impulse frequency. A further analysis could, however, be made
by using an integrating device kindly placed at our disposal by Pro-
fessor K. Granit. The amplified action potentials were then directly
recorded by one ray of the oscillograph through one channel, while
the other ray was driven by the integrator.

As the diaphragm was generally transected and the thorax thereby
opened while preparing the nerve, artificial respiration had to be
given by a Starling pump throughout the experiment. The arterial
blood pressure was recorded from the femoral artery by means of a
mercury manometer. Intravenous injections were made through the
femoral vein. In some experiments the cats were bled from the femoral
artery of the other leg. The blood was collected and heparinized for
reinjection. In some other experiments the sinus nerves on both sides
were exposed according to the method previously described (Gernandt
and Zotterman 1945).

EFFERENT IMPUT^ES IN THE SPLANCHNIC NERVE.

233

Besults.

Efferent impulses ivere regularly found in the splanchnic nerve,
though the intensity varied a great deal, which seems to be in
good agreement ^vith the well-known variations in the vascular
tone. Sometimes the action potentials were continuous, but more
often they occurred in groups, which might be synchronous with
respirations or vdth the heart beats. Similar observations have
already been described by Adrian, Bronk and Phillips (1932)
in different sympathetic nerv'cs suppl)nng constrictor impulses
to blood vessels, and by Bronk, Ferguson, Margaria and
SoLANDT (1936) in the cardio-sympathetic fibres.

The simplest way of increasing the efferent impulses in the
splanchnic nerve fibers is to stop the artificial respiration. The
nerve activity now rises gradually in jerky stages, followed by a
rise of the arterial blood pressure. After 45 to 60 seconds the
activity reaches its maximum, and the records show that a
synchronization of the fiber acthdty takes place (cp fig. 1 B).
MTien artificial respiration is now resumed, the potentials quickly
fade away, and after about 10 to 15 seconds there is a period of
nearly complete silence for about 10 seconds, after which, in
c. 30 seconds, the electric activity gradually returns to normal.
This silent state was observed when the blood pressure had begun
to fall but was still greatly elevated. A comparison vith the
noise level shows that the impulses were brought to a very low
level.

During asphyxia, spontaneous respiratory movements became
gradually stronger and sometimes so violent as to displace the
electrodes. We therefore thought it worth while to establish a
proper control, in order to ascertain whether the augmented and
synchronized potentials observed during the course of the asphyxia
were due to propagated efferent action potentials in the nerve
and not either to movements of the electrodes on the nerve or
to any secondary potentials in the nerve set up by muscular ac-
tion potentials of the violently contracting diaphragm or other
muscles over which the nerve passed. For that purpose we cura-
rized three of the cats used in these experiments. The result
was that the potentials appeared during asphyxia exactly as be-
fore curarization, although all muscular movements were com-
pletely inhibited.

234 B. QEENANDT, G. LILJESTRAND AKD Y. ZOTTERMAN.

ii'iWiAi «Hf>i>i*i»

AAAAAAAAAAAAAAAAAA

Fig. 1. Cat 4.3 kg, under chloralose anaesthesia. Action potentials from the
splanchnic nerve recorded directly (upper curve) and by means of an integrator
(lower curve). A. Standard ventilation air; JB. Asphyxia for I minute; G. Standard
ventilation air again for 10 seconds. D. Control. Nerve killed between the electrodes.

Time 50 cycles per second.

.00

160

1^0

KFFEUEN-T IMPULSES IX THE SPLAXCIIXIC XEllVE. '235

If artificial respiration was established with a gas mixture poor
in oxygen, the action potentials were greatly increased in magni-
tude (fig. 3), as was also the case when the air was replaced by
10 per cent carbon dioxide in oxygen (fig. 4 B). Thus both oxygen
want and carbon dio.xide accumulation increase the potentials.
On the other hand, if oxygen was given instead of air, a small

but definite decrease in the

activity was observed. This
must signify that, even with
air, a certain stimulation
from oxygen want occurs,
which enhances the impulses
in the splanchnic nerve.

Asphyxia could also be
obtained by bleeding the
animal, and the result was
a very considerable increase
in the impulses, though in
this case the blood pressure
was going down. It is striking
that the effect was observed
already 10 — 20 seconds after
the bleeding had started.

By retransfusion the picture
could be transformed to its
original state (fig. 5). A
similar result was obtained
when the blood pressure was
lowered by the injection of

acetylcholine (fig. 6 and 7), the change also appearing rather
quickly in this too. On the other hand, if the blood pressure was
raised by adrenaline (fig. 8), the impulses decreased and might
nearly disappear. These observations seem to indicate that the
effect of asphyxia is a combination of the stimulating action
of the oxygen want and the carbon dioxide accumulation, on
the one hand, and the inhibitory action of the elevated blood
pressure, on the other.

In order to get a better insight into the mechanism responsible
for the effects observed, the influence from the carotid sinus
and the depressor nerves was studied. Clamping of the carotids
gave an obvious increase in the potentials (fig. 9). As is well

30 *

T t

Fig. 2. Record of blood pressure during
nsphyxin from ■} to 1. Arrows A, B nnd
C refer to the momentB wlicn tlio records
A, B and C of fig. I were taken.

236

B. GERNANDT, G. LTLJESTRAND AXD T. 7.0TTERMAN.

AAAAAAAAAAAAAAAAAA

Fig. 3. The same cat as in previous figures. A . Standard ventilation air; B. Asphyxia
for 1 minute. 0. Standard ventilation air; D. Standard ventilation -with 7.3 % 0.

in N,.

efferent impulses in the splanchnic nerve,

237

*/'

^ r - — /'V-

(VV\AAAAAAAAAAAA/VV\I

Fig. 4. Cat, 2.8 kg. A. Standard ventilation air; B. Ditto with 10.5 % COj in 0.,
C. Standard ventilation air; D, Ditto with 100 % 0,.

238 B. GERN.ANDT, G. BIWESTRAND AND Y. ZOTTERMAN.

wwvwvvwwvvvwv

AAAAAAAAAAAAAAAAAA,

/WVV\AAAAAAAAAAAAA/\

'bin'll t iiif i'irii''iii i|ji Ai-~iir — <rn inmw>«\wi~''‘^~i"»iiiiin Wiinv i

wwwvvwwwwwv

Fig. o. Cat, 3.0 kg. A. Standard ventilation, BP 120 mm Hg; B. Bleeding from
femoral artery, BP 40 mm Hg; C. The lost blood was heparinized and injected
through the femoral vein, BP 140 mm Hg; D. Kepeated bleeding of o. 25 ml blood,

BP 40 mm Hg.

known from earlier experiments, this leads to a moderate rise in
the number of potentials in the sinus nerve that are elicited by
chemical stimulation (cp Euler, Liljestrand and Zotterman
1939), but a reduction of the potentials from the pressoreceptors.

239

EFFERENT IMPULSES IN THE SPLANCHNIC NERVE.

Iaaaaaaaaaaaaaaaaa/

vAAAAAAAAAAAAAAAAAA/

Eig, 6. Cat, 2.7 kg. A. Standard ventUation air BP 80 mm Hg; B. After injection
of 1.5 '/ acetyl choline, BP 50 mm Hg.

Both these effects contribute to raising the blood pressure (cp
Euler and Liljestrand 1943). The result obtained conforms
•well "with these findings, and demonstrates the influence of presso-
and chemoreflexes from the 'sinus mechanism on the actraty of
the splanchnic nerve under fairly physiological conditions.

Another proof of such an
influence can be obtained,
if the action potentials of the
splanchnic nerve are record-
ed before and immediately
after section of the sinus
nerve.s and the vagi. In order
to be able to compare the
results, it is of course neces-
sary to prepare the nerves
beforehand and cut them
without altering the position
of the electrodes. As is illus-
trated in fig. 10, the effect
of the operation was a great
increase in the amplitude.

Pig. 7. Record of blood pressure when
1.5 / acetyl choline was injected intra-
venously j. Arrows A and B refer to
records A and B in fig. 6.

240

B. QERIS'AKDT, G. LILTESTRAND AKD T. ZOTTERMAE.

AAAAAAAAAAAAAAAAA/i

. i

Fig. 8. Cat, 2.8 kg. A. Standard ventilation air, BP 150 mm Hg. B. After injec-
tion of 10 y adrenaline, BP 200 mm Hg.

WWWVXAAAAAAAAAA/

aaaaaAaaaAaaaaaaa/>

^ u r«iy» v

Fig. 9. Cat, 2.7 kg. A. Standard ventilation air, BP 90 mm Hg. B. Carotid arteries
clamped, BP 100 mm Hg.

EFFEKEKT iMPl5E=F^ IX THE SI-LAXCIIXIC XEltVE,

241

Fig. 10. Gat, 2.n kg. A. Standard ventilation, BP 105 mm Hg; B. A8[)liyxia for
85 seconds, BP 200 mm Hg; O. Standard ventilation after tlie soverahee of all
buffer nerves, BP 105 ram H g; D. Buring aspby.xin for 25 seconds, BP 1 1 5 inro Hg.

though the blood pressure in the experiment in question remained
constant. There was still a further increase during asphyxia, as
shown in fig. 10 D. But neither pure oxygen nor oxygen want
had now any influence on the magnitude of the potentials, though
a small increase was observed when 10 per cent carbon dioxide
16 — i6021S. Acta plij/s. Scandinav. Val.ll.

242

B. GBRNANDT, G. ULJESTRAND AND T. ZOTTERMAN.

“ /Vr^Vs''

J\fV

^AAAAAAAAAAAAAAAAA/'

Pig. 11. Cat, 2.8 kg. After the severance of Hering’s nerve and the vagus on both
sides. A. Standard ventilation air, BP 240 mm Hg; B. Ditto -with 7.3 % 081017,,
BP 130 mm Hg; C. Ditto with 10.5 % CO, in 0„ BP 150 mm Hg; D. Ditto with
100 % 0„ BP 200 mm Hg.

efferent impulses in

THE SPLANCHNIC NERVE.

•243

AAAAAAAAAAAAAAAiAAA/

Fig. 12. Cat, 2.8 kg. Buffer nerves cut. A. Standard ventilation air, BP 140 mm
Hg. B. Ditto after 10 ■/ adrenaline intravenously, BP 240 mm Hg.

in, oxygen was given, (fig. 11). The effect of aspliyxia after j de-
nervation thus seems to be due entirely to the accumulation ol
fcarbon dioxide. If the blood pressure was raised by adrenaline
from 140 to 240 ram, no effect on the potentials was observed
(fig. 12). In another case, however, a considerable decrease in the
amplitudes was found after denervation, when the blood pressure
after adrenahne rose from 50 to 100 mm.

Discussion.

The results obtained clearly show that the impulses in the
splanchnic nerves ate greatly modified by influences over the
sinus and aortic mechanisms. These effects arc exercised by altera-
tions in the chemical composition of the blood as well as by varia-
tions in the blood pressure. In our experiments oxygen want
was found to cause a very considerable increase of the impulses
occurring during air-breathing, whereas a diminution was observed
when oxygen was substituted for air. These effects disappeared,
however, when the buffer nerves had been eliminated. The sti-

244 B. GERNANDT, G. LIUESTRAND AKD Y . ZOTTERMAN.

mulating action of oxygen want on the vascular center is there-
fore mostly indirect. This corresponds well with the effects of
alterations in the oxygen pressure on the activity of the sinus
nerve (Euler, Liljestrand and Zotterman 1939) and also on
respiration, blood pressure and heart rate (cp Euler and Lilje-
STRAND 1942), though the results on circulation may be more or
less obscured by the regulation via pressoreceptors. On the other
hand, in our experiments, carbon dioxide stimulated the action
potentials in the splanchnic nerve even after denervation, though
probably to a smaller degree than before. This would imply that
the action is exercised both reflexly and directly on the center.
This seems to be in harmony with known facts about the influ-
ence of carbon dioxide accumulation on the sinus nerve activity
and on respiration and blood pressure. It must be pointed out,
however, that Alexander (1945) observed an increase in the
action potentials in the inferior cardiac nerve during oxygen
want, even after isolation of the upper thoracic cord from all
other nervous influences by low- and mid-thoracic transections
of the spinal cord and section of all dorsal roots and of the sym-
pathetic chains between those transections. The author himself
points out that the possibility cannot be excluded that the activity
in the inferior cardiac nerve in the deafferentiated spinal prepara-
tion might be due to irritative effects from the trauma of the cord.

With regard to the results observed by us during asphyxia,
it is obvious that the stopping of the artificial respiration will
lead to oxygen want and carbon dioxide accumulation, both of
which will act on the sinus and aorta mechanisms, evoking im-
pulses that stimulate the vasoconstrictor center. Carbon dioxide
will also act on the center directly. As a consequence of these
influences, the blood pressure rises, and now extra inhibitory
impulses are elicited by the stimulation of the pressoreceptors,
and these impulses work in a direction opposite to that of the
chemical stimulation. The net result will be determined by the
difference in effect between the two sets of impulses. When arti-
ficial respiration is reestablished, the chemical stimulation quickly
disappears, but the blood pressure may still be elevated — a
similar prolonged elevation of blood pressure due to inertia of the
effector system has been described by Pitts, Larrabee and
Bronx (1941) — and inhibitory impulses will then remain strong.
This seems partly to explain the fact that there is a short period
just after the blood pressure is beginning to fall when the impulses

EFFERENT IMPULSES IN THE SPLANCHNIC NERVE. 245

are less frequent than before asphyxia or even disappear entirely.
Since Pitts, Labrabee and Bronk have been able to demonstrate
that increased discharge in the inferior cardiac nerve during
hypothalamic stimulation is followed by an inhibition even after
section of the buffer nerves, it seems very probable that a re-
duced excitability of the sympathetic center follows intense ac-
tivity. The above-mentioned after-effect when artificial respira-
tion is started again might therefore be caused to some extent
by -the result of this lowering of the excitability.

According to the experience of the school of Heymans (cp
Heymans, Bouckaert and Regniers, p. 110), the vasomotor
centers and the centers for the adrenaline secretion are them-
selves insensitive to physiological variations in the arterial pressure,
whereas extreme reduction of the pressure may exercise a'directly
stimulating influence. This last effect must certainly be ascribed
to chemical stimulation. When the blood pressure had been
raised considerably from the normal level by adrenaline, we found
that the activity of the splanchnic nerve disappeared more or
less. This did not happen when the buffer nerves had been cut.
The conclusion must be that the great reduction of the potentials
in the splanchnic nerve is the result of a very strong inhibition
from the pressoreceptors. But this is not the only effect. A cer-
tain reduction of the normal chemical stimulation is also to be
expected. This would be in harmony with the corresponding
immediate effect on respiration of adrenaline, involving a re-
duction that may even lead to apnoea. Most of this effect on
respiration is due to the abolition of the chemical stimulation
from the sinus and aortic bodies (cp Gebnandt, Liljestrand
and ZoTTERMAN 1945), but a small part of the effect can be ob-
tained after section of the buffer nerves and may be attributed
to an improved circulation through the center, which will dimin-
ish the accumulation of carbon dioxide. A similar influence of
adrenaline on the vasomotor center seems probable. This will
especially be the case if the circulation is inadequate. Thus,
when the blood pressure was very low, adrenaline was found
greatly to inhibit the impulses after denervation of the sinuses
and section of the vago-depressor nerves. Some small effect might
also be expected when the circulation is adequate, and we note
in this connection that after the elimination of the buffer nerves
slight inhibiting effects of adrenaline on the ' action potentials
of different sympathetic nerves have been observed by Adrian,

246 B. GERNANDT, G. LILJESTRAND AND Y. ZOXTERMAN.

Bronk and Phillips ( 1932 ), Pitts, Larrabee and Bronk
( 1941 ), and by Alexander ( 1945 ).

The comparatively great effect of bleeding on the action poten-
tials of the splanchnic nerve is explained by the assumption that
in this case asphyxia is accompanied by a lowering of the blood
pressure, which in itself will reduce the inhibitory influence from
the pressoreceptors. The same holds true for the effect of acetyl-
choline. Probably in both cases local asphyxia sets in rather
quickly, which would explain the rapid development.

The assumption that the buffer nerves influence vasodilators
as well as vasoconstrictors, would imply that, corresponding to
a diminution of the efferent pressor impulses in the splanchnic
nerve, and increase in the dilatory impulses appears. Thus, e. g.,
at the height of adrenaline action or shortly after asphyxia, one
might expect such impulses. We have been unable, however,
to observe anything that can be attributed to such an activity.
Of course there is the possibility that the fibers concerned might
be much smaller than those leading to vasoconstriction. They
might thus escape observations.

Summary.

Efferent impulses were observed in the splanchnic nerve of the
cat, sometimes continuous, sometimes synchronous with respira-
tion or with the heart beats.

During asphyxia the amplitude of the impulses increased greatly.
If artificial respiration was resumed, the electric activity quickly
diminished and for a while nearly disappeared. Oxygen want
led to an increase in the potentials, and pure oxygen to a de-
crease. Both these effects disappeared after the elimination of the
buffer nerves. Accumulation of carbon dioxide caused greater
activity in the nerve before as well as after section of the sinus
and vagodepressor nerves.

Bleeding, or the injection of acetylcholine, led to a remarkable
increase in the potentials, whereas adrenaline gave rise to a di-
minution or even the disappearance of the impulses. After de-
nervation of the sinus and aorta mechanisms, the effect of adre-
naline was very small at normal blood pressure but fairly great
at low level.

Section of the sinus and vagodepressor nerves greatly increased
the electric activity of the splanchnic nerve.

EFFERENT IMPULSES IN THE SPLANCHNIC NERVE.

247

The results are interpreted as being due to reflex influence from
the chemo- and pressoreceptors of the sinus- and aorta mechanisms
and to a central effect from carbon dioxide accumulation.

This investigation has been aided by grants from the Therese
and Johan Andersson Memorial Foundation.

References.

Adrian, E. D., D. W. Bronx, and G. Phillips, J. Physiol. 1932.
74. 115.

Alexander, R. S., Amer. J. Physiol. 1945. 143. 698.

Bacq, Z. M., F. Bremer, L. Brouha, and C. Heymans, Arch. int.
Pharmacodyn. 1939. 62. 460.

Bacq, Z. M., L. Brouha, and C. Heymans, Ibidem 1934. 48. 429.
Bernthal, T., H. E. Motley, F. J. Sch-wind, and W. F. Weeks,
Amer. J. Physiol. 1945. 143. 220.

Bernthal, T., and F. J. Schwind, Ibidem 1945. 143. 361.

Bronx, D. W., L. K. Ferguson, R. Margaria, and D. Y. Solandt,
Ibidem 1936. 117. 237.

Cyon, E., and C. Ludwig, Ber. sachs. Ges. d. Wiss. math.-phys. Kl.
1866, 18. 307.

Euler, U. S. v. and G. Liljestrand, Skand. Arch. Physiol. 1934.
71. 73.

Euler, U. S. v., and G. Liljestrand, Acta Physiol. Scand. 1942. 4.
34.

Euler, U. S. v., and G. Liljestrand, Ibidem 1943. 6. 319.

Euler, U. S. v., G. Liljestrand, and Y. Zotterman, Skand. Arch,
Physiol. 1939. 83. 132.

Gernandt, B., G. Liljestrand, and Y. Zotterman, Acta Physiol.
Scand. 1945. 9. 367.

Gernandt, B., and Y. Zotterman, Ibidem 1945. 9. 362.

Gernandt, B., and Y. Zotterman, Ibidem 1946. 11.

Heymans, C., J. J. Boucxaert, U. S. v. Euler, and L. Dautrebande,
Arch. int. Pharmacodyn. 1932. 43. 86.

Heymans, C., J. J. Boucxaert, and P. Regniers, Le sinus carotidien,
Paris 1933. 334 p.

Jansen, W. H., W. Tams, and H. Achelis, D. Arch. Min. Med. 1924
144. 1.

Kramer, M., and S. Wright, Quart. J. exp. Physiol. 1932. 21. 319.
Ludwig, C., and L. Thiry, Sitz.-ber. Ak. d. Wiss. Math.-nat. 01
Wien 1864. 49 (2). 421.

Pitts, R. F., M. 6 . Larrabee, and D. W. Bronx, Amer. J. Physiol
1941. 134. 359.

Schneider, D., Arch. exp. Path. Pharmak. 1934. 176. 111.
Zotterman, Y., Skand. Arch. Physiol. 1936. 75. 105.

From the Physiological Department, Karolinsha Institutet,
Stockholm.

The Ejffect of Respiratory Changes upon
tlie Spontaneous Injury Discharge of Afferent
Slaminalian and Human Nerve Fibres.

B. GERNANDT and Y. ZOTTERMAN.

Received 2 January 1946.

This paper deals tvith a spontaneous activity of nerve fibres
which sometimes occurs in splanchnic nerve preparations from
the cat. At first this phenomenon was a great hindrance only
to the fulfilment of our original plan of work. It soon became
clear, however, that the impulses were being discharged in afferent
fibres. The further study of this spontaneous activity gave an
explanation of the well known phenomenon of pcst-ischeniic
sensory tingUng in the human subject after renewal of the blood
flow- to a previously compressed limb. It was also shown that
carbon dioxide produced a great reduction of this tingling thus
confirming an old suggestion made by one of us (Zotterman"
1933), that the post-ischemic tingling was related to a lowering
of the concentration of metabohtes.

Spontaneous Impulses in Aflferent Fibres
of tlie Splanclinic Nerve of the Cat.

In the course of our research on the efferent impulse-traffic
in the splanchnic nerve of the cat (Gbrnandt, Liljestbastd and
Zotterman 1946) we sometimes observed a more or less pro-
nounced activity of the nerve consisting of fairly large spikes.
These spikes occurred in a very strong continuous flow, in con-
trast to the more jerky outflow of irregularly shaped action-
potentials of low amplitude which is characteristic of the effe-
rent discharge of this nerve. The general shape of the fairly large

EFFECT OP KESPIKATORY CHANGES.

249

spikes indicated tkat they must derive from fibres of a diameter
above 4 //. A microscopic examination of the splanchnic nerve of
the cat reveals that this nerve contains an abundance of mye-
linated fibres which can be divided in two main groups: a) fibres
■with a diameter of from 4 to 7 /,t; and b) fibres of a diameter of
from 1 to 3 The preganglionic efferent fibres belong to the latter
group, while the group of thicker fibres obviously consists of
afferent fibres conveying impulses from the Pacinian corpuscles
in the mesentery and noxious impulses from the abdominal
region. Judging from the amplitude of the action-potentials from
single fibres set up by mechanical stimuli applied to the Pacinian
corpuscles of the mesentery, the largest fibres of the splanchnic
nerve must be considered to convey impulses from these recep-
tors. Noxious stimuli most probably set up impulses in mye-
linated fibres of various sizes ranging from 6 to 1 as well as in
unmyelinated fibres. According to the nomenclature formerly
used by one of us (Zotterman 1939), impulses from the Paci-
nian corpuscles are conveyed in ^-fibres, while pain is conducted
by fibres belonging to the and d®-groups as well as to class C.

As has already been observed by Adrian (1930), the sponta-
neous impulses set up at the cut end of nerve fibres disappear
when the thick sheath is pulled off. He also states that when the
sheath is very thin (which is the case Avith the smallest nerves)
the action is as a rule very slight. We have been able to confirm
these observations in numerous cutaneous and other nerve pre-
parations.

Our first impression of this continuous flow of action-potentials
in the splanchnic nerve was that they should be looked upon as
injury potentials set up at the cut peripheral end of the nerve.
The direction of the monophasic response, however, indicated
that the impulses were conducted in a centrifugal direction.
The nerve was prepared and the electric activity was recorded
in the same way as was described in our previous paper (Gernandt,
Liljestrand and Zotterman 1946).

By moving the electrodes 20 — 30 mm along the exposed nerve,
which is suspended in the air, we were able to convince ourselves
that the discharge was not set up in that part of the nerve, but more
centrally. In another preparation which showed the same phenom-
enon we were able to disclose the fact that the discharge originated
from the part of the nerve where it left the tissues. The sheath
was drawn off to this point. This procedure of drawing off the

250

B. GERNANDT AND Y. ZOTTERMAN.

sheath from the cut peripheral end in a central direction is very
■easy for the first 10 to 20 mm. After that we gradually had to
pull harder, and it seems very likely that the nerve may be some-
what damaged just at the point where it leaves the surrounding
tissue. In following preparations which displayed this continuous
discharge of fairly large spikes, we were in any case able to de-
monstrate that the discharge took place from that part of the
nerve which was covered by the sheath and which was situated
very close to the point where the nerve entered the tissues. After
having dissected out the nerve further centrally, both electrodes
could be placed upon the part of nerve which was covered by
the sheath. Now the recorded direction of the spikes showed
that in this part of the nerve they were conducted in a centri-
petal direction.

It was thus absolutely proved that the impulses were not set
up in the naked and exposed part of the nerve. We stress this
-circumstance because such a finding would be incompatible with
the fact that the continuous discharge can be modulated by
■changes in the respiration of the cat. Thus we first observed that
the discharge invariably diminished in strength when the tracheal
tube was clamped. A further analysis revealed the fact that the
spontaneous discharge was very highly augmented by hyper-
ventilation, while hypoventilation diminished the activity. As is
shown in fig. 1, the spontaneous discharge of large spikes can
be very highly reduced by stopping the artificial respiration for
■30 seconds (fig. 1 c). In this case the great reduction in the dis-
charge of large spikes reveals the activity of the small pregang-
lionic fibres. 10 to 15 seconds after artificial respiration had been
started again this efferent discharge was quite abolished, as has
been previously reported (Gebnandt, Liljesteand and Zotteb-
MAN: 1946), The antidromic afferent discharge gradually returns,
however, and generally reaches its previous level in about one
minute (see fig. 1 d and e).

An interesting point was now to see how the antidromic dis-
charge reacted to anoxemia and to hypercapnia. The cat was
thus ventilated with 5.9 % oxygen in nitrogen. This gave a very
definite increase in the discharge (see fig. 2 b), while ventilation
with 13 % CO 2 in oxygen quickly reduced the activity and usu-
ally abolished all activity of this kind within 10 to 20 seconds
(see fig. 2 c). The irregularly shaped potentials seen in fig. 2 c
are due to the efferent preganglionic fibres.

KFFBCT OF KESPIRATORY GUANOES.

251

rwvwwwwwwvwwv

A/WWVVVWWVVVWVW\A

IAAAAAAAAAAAAAAAAAAAA^

c • ' '

iAAAAAAAAAAAAAAAAAAAA/\

lAAAAAAAAAAAAAAAAAAAAA/

i . ' .

Fig. 1. Records showing the continuous discharge of antidromic afferent impulses
in the splanchnic nerve of the cat. Chloralose anaesthesia 0.05 g per kg body- weight.

A. Standard ventilation of the lungs; B. after one minute of over- ventilation;
C. artificial respiration stopped for .“lO seconds; D. Artificial respiration resumed
for 10 seconds; E d:o 90 seconds later. Time marker 50 cycles ner second.

Fig. 2. Injury potentials from the splanchnic nerve of the cat.

A. Under standard ventilation with air; B. with 5.9 % O.inNj; C. with 13 %
CO,, which inhibits the discharge of large spikes and thus discloses the efferent
volleys of impulses in the preganglionic fibres (Class B fibres).

For the further analysis we considered it worth while to test
the effect of eliminating the effect of the carotid sinus and the
depressor activity by cutting Hering’s nerve and vagus on both
sides. This procedure had no influence upon the large spike ac-
tivity under standard ventilation, as will be seen in fig. 3 a.
This record shows that the preganglionic fibre activity is defin-
itely augmented, owing to the elimination of the inhibitory affe-
rent volleys from Hering’s nerves and the depressor nerves.
Ventilation of the lungs with 13 % COa in Oj has the same action
as previously upon the discharge of the large spikes, which is
reduced almost to nil, while the effect upon the preganglionic
fibres is now definitely increased (fig. 3 b). The effect of 5.9 %

KFFECT OF ilE.'PIUATORy ClfANGES.

253

oVVVXAAAAAAAAAAAAAAA.

AAAAAAAAAAAAAAAAAAA/

VWVVWWWWWVVW"

WVWWVAA/VVWWWW

Fig. 3. Records from the splanclinic nerve of the cat; same preparation ns in
figs. 1 and 2. Horing’s nerve and vagus out on both sides, whicli augments the

B-fibre activity.

A. during standard ventilation; B. artificial ventilation stopped for 30 seconds;
C. standard ventilation again; £>. artificial ventilation with 5.1) % 0, in Nj.

oxygen in nitrogen, however, had now disappeared. This can be
explained in the following way. As long as the chemoceptors
are intact the lotv oxygen pressure causes an increased ventilation
which results in a hypocapnia. After the cutting of all chomo-

254

B. GERNANDT AHD Y. ZOTTERMAN.

ceptive nerves low oxygen pressure does not induce any increase
in the ventilation, thus keeping up the COj tension as before
(see fig. 3 C).

Experiment on Man.

The effect of respiratory changes upon the spontaneous dis-
charge of afferent injury potentials in the cat induced us to study
similar phenomena in man. A similar spontaneous injury dis-
charge can easily be produced by pressure and asphyxia to the
nerves of the arm by inflating a cuff above the elbow and keeping
the pressure above the systolic blood pressure for a period of
5 to 10 minutes. The sensation of tingling following upon the
release of pressure has been thoroughly described by Lewis,
Pickering and Eotsohild (1931) and Zotteeman (1933). .Ku-
GELBERG (1944) has shown that the rheobase of the motor nerves
of the arm was essentially lower in tetany and during hyperven-
tilation of the subject. He has also shown that the slope of accom-
modation increases progressively as the ischemia goes on. And
finally, he has shown that after the release of the blood-flow to
the arm there is a sudden heavy fall in the slope of accommoda-
tion, and for a period of 2 to 7 minutes after the release there is a
break-down of accommodation. As far as we can see, the duration
of this break-down of accommodation coincides very well with
the temporary course of the post-ischemic pricking sensations.

Production of ‘post-ischemic tingling. The intensity of ting-
ling depends upon the time the arm has been compressed. As
shown by Zottebman (1933), the minimum compression time
varies individually but does not seem to extend over 7 minutes.
For this reason we have in our experiments chosen 7 minutes as
the standard time of compression. The cuff used was 12 cm broad
and was placed immediately above the elbow. It was distended
to 150 mm Hg. After release there is always a feeling of a wave
of warmness and heat in the skin below the cuff and then after a
silent period come the pricking sensations, which start in one or
more finger-tips, usually the thumb and index finger. The interval
between the moment of decompression and the start of the prick-
ing sensations varies individually and also from time to time,
as is shown in table 1, which gives the latency, the maximal
subjective strength and the duration of the tingling in two series
of experiments upon four healthy subjects. The subjects reported
the start and end of the tingling and gave the course of increase

EFFECT OF RESPIRATORY CHANGES.

25&

and decline of the subjective sensations. The intensity of the sensa-
tions was given in seven degrees, described as a gliding subjective
scale of sensations from the faintest pricking to a very painful sen-
sation, which is clearly reflected in the behaviour of the subject.

■\Vhen the subjects breathed air normally the latencies of
tingling varied from 55 to 80 seconds and the duration of the
tingling varied from 2 to 3 minutes. The maximum subjective
strength of tingling varied from 2 to 3 degrees, which means that
the tinghng at its maximum was quite distinct but not in any
way disagreeable.

As will be seen from table 1, the strong over-ventilation which
commences at the moment of decompression brings about a very
great change in the picture. The latency was now as a rule dis-
tinctly shortened, in one case to less than one half. The tingling
rose steeply to a very high degree and was reported as very painful
and disagreeable. In some subjects the over-ventilation led to
wide-spread symptoms. As long as the subject was over-ventilated,
the highly painful tingling went on without any sign of decrease.
For this reason the experiment was discontinued after 6 minutes.

In the light of our observations on the spontaneous discharge
of afferent impulses in the cat’s splanchnic nerve, we expected
that the phenomenon would be changed in the opposite direc-
tion by inhalation of carbon dioxide. This expectation was con-
firmed. Breathing from a Douglas bag a gas mixture consisting
of 7.1 % COj in air lengthened the latency of tingling. In one case
no tingling appeared. The maximum subjective strength as well
as the duration of the tingling was very markedly reduced. An
increase to 10.4 % of the COj-content of the inhaled air reduced
the tingling still more. This percentage of COo was of course
close to the upper concentration which can be endured for a
period of up to 5 minutes. The subjects now reported that the
tingling, which started after a still longer latency than before,
was so faint that they would scarcely have observed it if their
attention had not been especially directed thereto. The pricking
was felt only in one fingertip and on one spot only, and it was
experienced as a very faint pricking occurring with silent periods
up to 10 seconds.

An attempt to illustrate the strength and the course of the
post-ischemic tingling under the various conditions described
above has been made in fig. 4, the diagram of which is constructed
upon the data obtained from one of the subjects. (0. N.).

Table 1

EFFECT- OF IVESPinATORT CHANGES.

257

Bubjedivc
sirengfh of

Pig. 4. Diagram showing the latency and course of tingling after 7 minutes of
compression of the arm above tho elbow.

normal breathing of air; forced breathing of air;

— — — breathing of 7.1 % CO. in air; breathing of 10.1 %

CO, in air.

Discussion.

It has previously been suggested (Zottebi^ian 1933) that a
lowering of the concentration of metabolites in the tissues plays
a decisive role in the establishment of postischemic tingling. The
long latency for the tingling indicates that it is not elicited by
the mechanical action of the blood-flow forcing its way through
the previously compressed part, but shows that the tingling does
not set in until after the blood supply has been restored for half
a minute or more. The reactive hyperemia developed in the de-
compressed region no doubt is followed by a very ample blood-
flow, which quickly changes the reactive properties of the larger
myelinated nerve fibres in the direction shown by Kurelbeeg
(1944). It is thus not surprising to find that changes in the CO.-
tension of the blood will produce very definite quantitative
changes in the postischemic tingling sensation. These changes
in the tingling are obviously quite comparable to the spontaneous
discharge of injury action-potentials from afferent fibres observed
in the splanchnic nerve of the cat. Both phenomena are reduced
in strength or inhibited by an excess of carbon dioxide in the
circulating blood; on the other hand, the activity is strengthened
or initiated by a reduction of the carbon-dioxide-tension of the
blood.

17 — 't60215. Acta plnjs. Scandhiav. Vol.ll.

258

B. GBRlTAinJT AHB T. ZOTTEEMAN.

The changes in the carbon-dioxide-tension of the blood in these
experiments undoubtedly brought about very definite changes
in the pH of the blood and the tissues. Thus, breathing of 10 per
cent carbon dioxide will cause a definite increase in the amount
of ionized calcium in the tissues, while an increase of the pH
caused by strong over-ventilation of the lungs brings about a
lowering of the Ca-ion concentration.

Lehmann (1937) showed that readmission of oxygen to pre-
viously asphyxiated peripheral mammalian nerves caused within
1 to 2 minutes a rapid fall in the threshold to a normal value.
The threshold then rose for a second time and remained high
for about 50 minutes. This latter change in the excitability is
very similar to that brought about by a change of the pH ftom
8 to 7.4 in the surrounding solution. He also showed a further
important fact, i. e. that during the period of low thresholds the
afterpotentials are decreased, and the nerve becomes spontane-
ously active as it does in a state of low calcium.

Another point which is of interest in this connection is that
strong tactile stimuli to the finger-tips applied when the pricking
paresthesia is at its height cause a very sharp and severe pain.
It has been suggested by Zotterman (1933) that this phenomenon
is most probably caused by the pain fibres in the previously
compressed part of the nerve being stimulated secondarily by
the action-potentials of adjacent tactile fibres. Gbanit, Leksell
and Skoglund (1944) have recently demonstrated that an inter-
action of different fibres actually does take place in injured or
compressed regions of a nerve. Thus the increase of the pricking
paresthesia induced by peripheral tactile stimuli can be inter-
preted in the light of these facts. In a second paper Gbamt and
Skoglitnd (1945) state that the “artificial synapse” formed by
the cut end of a mammalian nerve is best demonstrated in de-
cerebrated cats which have not lost too much blood and in
cats under chloralose narcosis, while cats under “dial” were too
deeply narcotized. We- would suggest that this difference is not
effected by any direct action of the narcotics upon the peripheral
nerve, but depends upon differences in the ventilation of the
lungs^'

EFFECT OP KESPIRATORY CHANGES.

259

Summary.

1. Injury potentials set up in the afferent fibres of the splanch-
nic nerve of the cat have been shown to respond in a regular
manner to changes in the carbon-dioxide-tension of the blood.
Artificial over-ventilation with air as well as the increased ventila-
tion ensuing from breathing gas mixtures low in oxygen produce
a very marked increase of the injury potentials, while a rise in
the carbon-dioxide-tension of the blood inhibits the discharge.

2. In full accordance with the above phenomena it was found
that the pricking paresthesias following upon the release of the
bloodstream to the arm after a previous period of asphyxiation
can be modulated in a corresponding way by changing the carbon-
dioxide-tension of the subject; h 3 rpocapnia increasing them and
hypercapnia causing a reduction or abolition of the sensations
experienced.

3. These findings are discussed and related to the theory ad-
vanced by Lehrlann to the effect that the variations in the irri-
tability of mammalian nerve fibres produced by changes in the
Ch of the tissue are due to changes in the amoimt of ionized
calcium.

This work has been aided by a grant from the Therese and
Johan Andersson Memorial Foundation.

Eeferences.

Adrian, E. D., Proc. Roy. Soc. B, 1930. 106. 596.

Gernandt, B., 6 . Liljestrand, and Y. Zotterman, Acta Physiol.
Scand. (in press).

Granit, R., L. Leksell, and C. R. Skoglund, Brain. 1944:. 67. 125.
Granit, R., C. R. Skoglund, J. Physiol. 1945. 103. 435.

Kugelberg, E., Acta Physiol. Scand. 1944. Suppl. 24.

Lehmann, J. E., Amer. J. Physiol., 1937, 118. 600.

Lehmann, J. E., Ibidem 1937. 118. 613.

Lehmann, J. E., Ibidem 1937. 119. 111.

Lewis, T., G. W. Pickering, and P. Rotschild, Heart. 1931. 16. 1.
Zotterman, Y., Acta Med. Scand. 1933. 40. 185.

Zotterman, Y., J. Physiol. 1939. 95. 1.

Methods for Continuous Tissue Culture as
Applied to Bone Marrow.

CLAUS MONK PLUM.

Received 8 January 1946.

Experiments on the development of cells in vitro can provide
a certain amount of information about the development of vari-
ous tissues, their metabolic processes etc.

In hematological work tissue culture methods have up to now
only been used for isolated observations in connection with theo-
retical considerations, as for instance the denucleation of erythro-
cytes, the formation of thrombocytes etc., but it ought to be
possible to utilize them to approach the solution of numerous
hematological problems, and I have attempted to work out suit-
able methods for such purposes.

I shall in the following briefly describe an apparatus which
has been successfully applied for cultivation of bone marrow
cells over long periods of time and which will, I believe, be applic-
able also to other branches of tissue cultivation.

In 1936 (at about the time when the first attempts on culti-
vation of human bone marrow were made) Osgood pubhshed a
method for the culture and study of liquid bone marrow. His
intention was to provide conditions for the bone marrow suspen-
sion approaching the normal as closely as possible viz., the ap-
paratus would function as lung, kidney and organ of circulation
for the marrow, and at the same time the construction would
allow sterile samples of the suspension to be taken at any time
during the progress of an experiment. Simultaneously Osgood
made a number of suggestions concerning inquiries into condi-
tions of cultivation and development of cells for which the appa-
ratus would appear to be suitable in view of the physiological
conditions aimed at. In spite of these possibilities Osgood and

METHODS FOR CONTINUOUS TISSUE CULTURE. 261

his coworlces replaced the method by a simpler one after half
a year’s trial. In tliis new method the close approach to physio-
logical conditions was abandoned and a constant composition
was not maintained neither for the ventilating air current nor
for the culture medium.

Although Osgood’s second method would be easier to work
with, I considered the physiological conditions so important that
I decided to base my construction as far as possible on the prin-
ciples of Osgood’s first method. I found it necessary to work
out two constructions, one for the prolonged experimentation
on fairly large samples and another, embodying as far as possible
the same principles, but allowing continuous microscopic observa-
tion of the cells under culture.

1. A method for the cultivation of large samples.'

After a number of preliminary experiments with an apparatus
very similar to Osgood’s first construction I adopted a some-
what modified arrangement which, while still fulfilling the pri-
mary conditions, was in many ways easier to work with.

For practical reasons I had four identical apparatus constructed
and all of them mounted in the same thermostatic water bath,
a point of considerable importance for quantitative comparative
work.

As stated above normal physiological conditions were the main
consideration, and it is possible to regard the apparatus as made
up in sections representing respectively the supply of nourish-
ment, the respiration, the circulation and the excretion of waste
products. These sections are to a certain extent separated by
membranes permeable to gases and crystalloids and correspond-
ing more or less to the vascular endothelium and the kidney
glomerular membrane. In the original Osgood apparatus “parlo-
dion” was used for these membranes, but owing to the war this
substance was unobtainable. I first prepared membranes from
collodium, which were difficult to get uniform, and later I used
the ultrafiltration membranes described by Kehberg (1943) and
prepared by the Copenhagen firm “Kapcello”.® These were made

' The macro-apparatus avas made by IJ. Slmers chemiske Lahoratorhim, Copen-
hagen, Tvho will be ready to supply it under my supervision. I want to thank Mr.
Rasmussen, cand. phnrm,, of the firm for his helpfulness during the construction.

^ “ My thanks are due Mr. Hawlik, Civil Engineer of this firm for his kind as-
sistance in providing these necessar^^ mcmhrane.s.

262

CLAUS MUNK PLUM.

for me in two sizes, viz. diameter 18 mm, length 110 mm, and
diameter 9 mm, length 60 mm. Both are practically impermeable
to colloids.

I shall now describe in detail the apparatus as in use during
an experiment, referring by numbers to the figures 1 and 2.

Diagram of the apparatus for tho cultivation of largo samples of bono-marcow colls in suspension.

METHODS FOE CONTINUOUS TISSUE OULTUEE.

263

The provision of
nutriment.

In large scale tissue
culture work it is impor-
tant to provide a con-
tinuous supply of food for
the culture. This is done
in this apparatus by a
constant flow of the ap-
propriate solution through
the internal membrane
(5, fig. 2). The solution
enters through the tube
in the top left corner ( 1 , 1 )
passes through the T-tube
(2, 1 ), the drip vessel (3,1),
internal diameter: 18 mm,
and the capillary tube
(4 , 1 and 2 ), internal dia-
meter: 1, 2 mm, which
opens near the bottom of
the 9 mm membrane (5,2).
From- here the fluid rises
through the membrane
and the tube (6, 2 ), in-
ternal diameter: 7 mm,
to which the membrane
is attached and leaves
through the side tube
(7, i), where it can be
collected in a bottle (not
shown). The rate of the
flow of the fluid was in
my investigations 80 — 90
ml/hour.

Fig. 2. A fragment of the diagram fig. 1
showing the membranes placed in the
apparatus.

The nutritive substances diffuse out through the membrane
(5, 2 ) thereby supplying the culture in the membrane (8, 2 ) and
the concentration here will remain approximately constant when
the supply is large compared with the amounts used up.

264

CLAUS MUNK PLUM.

The membrane (8, 2) contaming the cell suspension — bone
marrow cells suspended in Ringer-solution — under culture is
fastened to the glass funnel (9, 2) resting on three projections in
ward from the outer vessel (10, 1 , 2 ).

Aeration and determination of respiratory
metabolism.

The cell suspension (in 8, 2) is kept in continuous movement by
a current of air (or other gas mixture) through the tube (11,
1 , 2 ) and thereby prevented from sedimentation. The rate of the
air flow was 100 ml/hour.

The air which is made CO^-free previously is let into the appa-
ratus through the tube (12, 1 ) and passes through the vessel
(13, 1 ) for condensing water vapour. (14, 1 ) is an arrangement
for taking of gas samples for analysis through the rubber by
means of a cannula and a common air recipient. Passing on to
the vessel (15, 1 ) inserted for the sampling of the cell suspension
(see later) the air is filtered through cotton wool and passes through
the tube (11, 1 , 2 ) to the membrane (8, 2) in which some O5 will
be absorbed and COj added. The (expired) air passes through
the tube (16, 1 ) first to the glass vessels (18, 1 ) for the condensa-
tion of water vapour and then to the COj-absorber (17, 1 ). In or-
dinary experiments the absorber contains lime-water [Ca(OH)2]
stained with phenolphthalein to estimate the COg production,
but in special respiratory experiments the absorber is replaced
by an almost vertical tube (75°) length 750 mm, diameter 12 mm
containing Ba(OH)2 and the air is admitted from below through
a capillary tube placed in a rubber stopper. This arrangement
was very effective for absorbing and titrating small quantities
of COj. The Ba(OH)„ was emptied, by refnoving the rubber stop-
per, down in a small flask and titrated (lYiiirKLER 1931). YTien
determinations of COj-production were made one of the four ap-
paratuses was used as a control and the membrane (8, 2) filled up
Avith Ringer-solution without cells. Some CO. will diffuse out
through the “glomerular” membrane (8, 2) into the outer vessel
(10, 1 , 2 ) and, to secure this, “inspired” air is added through (19
and 20, 1 ), bubbles through the solution for removal of waste
products and leaves the air from the cell suspension.

METHODS FOR CONTIHDOUS TISSUE CULTURE.

265

Removal of waste products.

For teclmical reasons it was not found practical to have a
continuous flow of solution along the oustside of the membrane
(8, 2). A large volume of solution (15 x the cell suspension) was
therefore placed between the membrane (8, 1, 2) and the outer
vessel (10, 1, 2) and changed at intervals which had to be fairly
short (30 min.). The change is performed as follows: The solu-
tion (in 10, 1, 2) is removed by clamping (26, 1) on the expired
air tube. This raises pressure and drives out the fluid through
(27, 1, 2), (28, 1) pro\’ided with an arrangement for samphng
(29, J) to the funnel (30, 1) and the eyhnder (31, 1), where it is
collected. The funnel (30, J) has a reservoir with alcoholic neutral
red for determination of the pH, when desired. The cylinder can
be emptied out after closure of (32, 1) by means of the pump
(33, 1).

Fresh solution (usually suitably buffered Ringer) of body
temperature is contained in the reservoir (21, J). ^Vhen the chp
(23, J) is opened the solution will flow through (24, J) and (26,
i, 2) and fill up the chamber (10, J, 2), emptied just previously.

It is of course necessary to work in sterile conditions. Since
the complicated apparatus would be difficult to sterilize com-
pletely I preferred to work antisepticaUy by adding “solbroF’
(methyli-para-oxybenzoas) in a concentration of 1 ®/co to all
solutions. This proved harmless for the cells studied.

Before use the membranes were kept in water with added
“solbrol”, the Ringer solution used for cell suspension and waste
elimination was sterilized by boiling and “solbrol" was added.
The bone marrow was sucked out through a sterile cannula into
a sterile syringe of 20 ml capacity. After this the cells (0.5 ml)
were suspended in 25 ml Ringer solution. The technique used
for getting the bone marrow cells is to be published in detail
elsewhere.

In the experiments so far made infections have occurred only
very rarely, but then of course spoiled the experiment.

An experiment is started as follows: The membranes are mounted
m the apparatus and tested for tightness by means of Ringer
solution, especially at their connection with the respective glass
tubes brought about by 5 mm broad rubber bands of 1 mm thick-
ness, the diameters are a little smaller than the respective glass
tubes. The whole apparatus is put together and air sent through

266

CliATJS MUNK PliUM.

for 15 minutes during -wldcli tlie solution in th.e outer chamber is
changed 3 to 4 times. Next the flow of air is stopped by closing
the clips (38, 1, 39, 1 and 26, 1), and the Ringer solution in the
membrane (8, 1, 2) is sucked up in the tube (15, 1) by means of
a cannula introduced through the vacuum-rubber tube (42, 1)
and a syringe, capacity 30 ml. Then another cannula is pierced
through the vacuum-rubber (40, T) and the Ringer solution is
sucked out. After removing the cannula in the rubber tube (40,
1), the sterile bone marrow suspension is injected into the tube
(15, 1) through the cannula in the rubber (42, 1) and the cannula
is removed. The volume of the bone marrow is then read by
means of the graduation in the tube (16, 1) diam. 8 mm and
graduated in 0.6 ml.

Then the clips (26, 1 and 38, 1) are opened and the air flow
will press down the bone marrow suspension into the membrane
(8, 2). When the suspension is down in the membrane the clip
(39, 1) is opened.

When samples of the suspension are to be taken during the
experiment it is sucked up into (16, 1) and the volume is measured,
then the necessary quantity is drawn out by means of a small
syringe — 2 ml — from (40, 1).

In cases in which a flow of nutritive solution is not desired,
the membrane (5, 2) may be compressed and to avoid this water
may be introduced into the chamber (41, I) opening into the tube
(6, I) from the reservoir (36, I) and thus raise the pressure.

In some cases it may be practical to arrange a flow of the nu-
tritive solution also through the compartment (10, 1, 2) and to
effect this the second branch of the T-tube (2, 1) is connected
through the tube (37, 1) also with (10, 1, 2) and the solution in
the chamber (10, I, 2) is changed as described.

The apparatus as here described may at first sight appear
more complicated than Osgood’s first model, but in return for
this the life conditions of the cells are sufficiently good to main-
tain the normal structure of the blood corpuscles and therefore,
probably, also their functions. It should be mentioned that while
in the first experiments, in an apparatus copied from Osgood,
destructive changes were often observed in the erythrocytes,
even a crenation was only very rarely seen in the apparatus de-
scribed.

The apparatus as described wiU render possible the study of
a number of hematological problems including metabolism de-

•adoosojconir lepun v uo psocid st 8ioq.M aqj,

‘(p ‘S' ‘ox) sdqo OAN.!^ iCq UAVop ppq ‘rata iT'Q— XT'O ssaa^ioni;^
‘(p 6) dqsiaAoo ^acaq)a:o we P^® aoisaadsns qgo oq;}

ox (p ‘8 — L) ^0X^00 o^XX oioq fnra g e qxX“. dqswAoo (rara ii'o
— ^I'O) tnqx JoqxotcB (p ‘g) aatjiqraora aorpopoo 8aTJ[d arqx "b si
areSt? {p ‘g) dqsiaAoo aqx OAoqy -aoixtips aqx jjo gntiCjico pat?
aoixaps aATXTiX®^a Sarpaaj ioj. vCpAixosdsai °aiAi9s ‘saqoaaix opF
c^sB[ OAS.X o^X ^ is ‘Q P^® f) poppoqoia oit? SB{naat?o Ojax uaqj, rao
pa]'B 9 s iAoa SI {p ‘s) rara g jo ajoq pax^oo pgpiAoad — Sui

-pnoo sapsndioo poop aoj quo pnsn gqx — ip ‘Z) dr|si9A0D t? jaq
-raiiqo aqx Jo dox aqx '“O '^oq^^qo Saixanoo aqx a9j ox paoB^d sum.
aoaBxsqns aratts aqx jo x^tmorac qt?ins e pat? xpq^st? qxpu. is ‘l)
dn papj saqoaaax opts oAix aqx g *°ij tit aAvoqs sy 'paifoidraa
seM. aaqraBqo Saixanoo qoinjQ-aaqjng e spx ^oj siSBq e sy -saoix
-onixsai Am. o;^ Sapiooot? ‘aaStjquadoQ ‘NasiaHoij^ pus sooag;
TOiij: aqx jCq ata aoj apcra sbaa ‘adoosoaoira aq:^ aapun paoBid aq
ppoo qoiqAL snc^BaBddB ub ‘spx op o:f auo apBua ox pUB ‘qao atp
-oixiBd B ai saSaBqo aqx av.ojioj o^ ajqB aq ox xtiaxtodrat aaxjd st xi
‘S[BAaaxai ^e paanoas sajdaiBS aAiqBxuasaadaa ao ipo papnxs aq aao
aoB[d SapBX saSuBqo aqx aAoqB paquosap sb snxBaBddB aqx aj

•ain^ino iAOJJtjm ouoq eqi^ aoj poiixara-ojoiui y 'n

•aanxino aqx p ifpnxs oidoosoaoira
paxdtiaaaxapxn ob SAVoqB paquosap aq ox Avon aoixBOijfipoiu oaQ

•rasiaBSao aqx uiqxtAt snoixpaoo
IBiaaoa iqBoiSojoiSiCqd aqx j^iasojo aaora qoBoaddB ox sb os paAoadrai
osjB puB asodand ^aa aoj paijipora iXpBxras aq aso xi ‘Aji^m
-raqaad sb paaapisaoo aq ox si paqiaosap aaaq sb anbmqoax aqj,

'pjo sanoq

g2, saanxpo uo saoixB°ixsaAxn apBoi ssq aq x^qx sxaodaa aooDSO
Xnq ‘sanoq fg aoj aanx^no aqx qdaq aABq j sxaararaadxa jfra aj
'oxa saxtixpo ptaaxoBq ao saoaBpqns pppiiaxonq jo xoa^a aqx no
saipnxs ‘sqao jo sdnoaS paxB[ 0 si jo pan AioaaBta anoq aqx jo rasqo
-qBxara XaoxBaidsaa aqx jo saopBtnraaapp ‘'oxa saaoraaoq ‘saiiuBxiA
‘spioB oaraiB SxtiatadojaAap pan aopBaaji]oad qao aqx aoj XaBSsaoaa
saoaBxsqns aqx jo aopBannaaxap ‘aanjpo qao aqx aoj aotsaax ®00
pan aaSiCxo ‘jjd ‘aanxBaadraax jBoiixdo aqx jo snoiXBOiraaaxap 'zia
‘ dn paaado saixqiqissod aqx jo araos azraBtnrans iljjaiaq jpqs j

-snaojo

sdnoas paxB[osi no ao apqAi v sb AioaaBra auoq aqx no suorxBniraaax
19Z '3U0Z1Q0 anssii sfloiiKiiiioo uoj scrouisre

268

CLAUS MUNK PLUM.

Fig. 3. Diagram of the apparatus for the cultivation of isolated cells
of the hone-marro'w.

An enlargement of the central part of the apparatus shovm in Fig. 3.

Nutritive solution, suitably beated and saturated •with oxygen,
flows from a Dewar vessel through, the apparatus at a rate of
80 ccm/hour (to keep up the temperature).

This apparatus may be constructed a little cheaper when not
using a Burcker-Tiirck counting chamber but by using a 3 mm
thick object slide as basis and then mount by means of heated
Canada balsam some glass pieces to form the lower drain made
by the trenches of the counting chamber.

The observation of the cells is rather difficult since they must
remain unstained and only differences in size and shape can be
observed. Usually, however, the nucleus appears a shade lighter
than the rest of the cell content. The use of Zeiss’ “Phasenkon-
trast-Einrichtung” kindly lent by the Broch and Michelsen
somewhat facihtated the observations.

The two types of apparatus described above have been used
since May 1944 and proved satisfactory although requiring a
certain amount of training. They have been used so far only for
investigations concerning the erythropoiesis and the main re-
sults, which will be published in detail elsewhere, are as follows;
A new formation of red cells takes place even with Binger solu-
tion as the feeding liquid. The numbers of mitoses observable in

HETUODS FOR CONTINUOUS TISSUE CULTURE. 269

differential counts appear able to account for only 1 per cent
of tins new formation which must therefore take place by a
different mechanism. When liver extract is added to the feeding
fluid the number of eryt}irocytc.s formed is increased in a certain
relation to the concentration of the extract, but without any
increase in the number of raito.<ics. When the mitotic activity is
completely paralyzed by the addition of colchicin to the feeding
liquid — as well as to the cell suspension — new red cells are
produced at the same rate a-s before.

It i.s to be concluded from this .study that erythropoiesis takes
place in two .separate .stage.s. In the first stage normoblasts are
fonned by mitotic cell division, while in the second immature and
small red corjiuscles arc formed from the normoblasts by segre-
gation of protopla.sma pseudopodia. The normoblasts finally die
after giving off a certain number (about 100) of corpuscles.

Lisa I3o.stro.m (1910) .suggc,stcd a formation of crjdhrocytes
from normoblasts by segregation of pscudopodia without any
nucleus, and by means of the micromethod described above I
have been able to observe the process as follows: the initially
spherical normoblast gradually (10 — ^15 min.) became oval when
fed ■with liver e.xtract, mo.st of the protoplasm flowed toward
one end opposite the nucleus and suddenly a drop of protoplasm
was cut off and a .small erythrocyte formed. This small erythro-
cyte slowly grew in .size and in one hour or so it got nearly the same
shape and size an the normal erythrocytes in the suspension.

The constructions of the apparatus described above were made
possible by a grant from the Cnrhbcrg FoitndaUon and the work with
them supported by a grant from thcAVng Christian X Foundation.

I^Iy sincere thanks are due to both these foundations. I want
to thank also the firm I\Iedicinalco Ltd for the opportunity to
carry on thi.s investigation in the biological laboratory.

Litorature.

Bostrom, L., Nord. Med. 1940. 22 . 805.

Osgood, E. E. and B. S. Sluscovm, J. Amcr. med. Ass. 1936. 106 . 1888.
Osgood, E. E. and I. E. Brownlee, Ibidem 1936. 107 . 123.
Eehberg, P. Brandt, Acta Physiol. Scand. 1943. 5 . 305.

Winkler, Kolthoff, Die Massanalyse II, Berlin 1931.

From the State Pharmaceutical Laboratory, Stockholm,

The Influence of Different Temperatures on
the Action of Drugs on Autonomic
Effector CeUs.

HAKA2T RYDIN.

Received 18 Jannary 1946.

Fundamental investigations by Elliot, Loewi, Dale, Cannon,
and others have given us a considerably greater knowledge than
before regarding the physiology of the autonomic nervous system.
In coimection with Dale’s theory of the chemical transmission
of nerve impulses, we have obtained a better explanation than
heretofore of the mechanism and site of action of autonomic
drugs, even if many problems of considerable importance in this
field are still unsolved.

Among methods for the interpretation of the pharmacological
action of drugs, in vitro experiments on isolated surviving organs
are employed. In such experiments it seems by no means incon-
ceivable that certain structures of the effector cells are more
sensitive to temperature changes than others. In that case, rais-
ing or lowering of the temperature as compared with the normal,
in different ways as regards drugs with different sites of action,
might affect the response of the effector organ.

The author of this paper has therefore studied the problem
whether the temperature could be made the basis of a test which,
in certain cases, might contribute to greater knowledge of the
mechanism of action and classification of autonomomimetric
drugs, or confirm results obtained with other methods.

To judge by the literature, this problem does not appear to
have been previously subjected to investigation. On the other

ACTION OF DRUGS ON AUTONOMIC EFFECTOR CELLS. 271

hand, isolated studies have been published, in another connec-
tion, regarding the bearing of the temperature on the action of
certain drugs on surviving organs.

Thus, Laubender and his co-workers, in a number of publications,
have reported their observations on the action of some drugs (hist-
amine, acethylcholine, orastin and coniin) on smooth and striated
muscles, with special regard to the relation of the effects to the con-
centration of the drug and the reaction temperature. Their studies
were made on the gastrocnemius of the frog at temperatures ranging
between 5 and 20° 0 and on the uterus of guinea pigs between 18°
and 37°. As a rule, a stronger effect was obtained with rising tempera-
ture except for acetylcholine, where the intestine showed a maximum
effect at about 28°.

Rentzhog (1930), in a study of the effect of epinephrine on isolated
surviving small intestine of rabbit at varying temperatures, found a
maximum effect at about 28°. A similar observation was made by
Blaschko and Schlossmann (1938), who observed that the small
intestine of rabbit was most sensitive to epinephrine at about 30°.

Zadina (1938) observed that histamine gave the strongest effect
on the isolated small intestine of guinea pigs at temperatures below
37°. Emmelin, Kahlson and Wicksell (1941), in regard to the intes-
tine of guinea pigs, found a temperature of 32° in the ‘bath fluid to be
the most favourable in the estimation of histamine in the plasma.

1. Methods.

The author’s experiments were performed on the isolated surviving
small intestine of rabbits according to the technique of Magnus.
The intestinal segments were suspended at 38° C in Thyrode solution,
consisting of NaCl 0.8 %, KCl 0.02 %, CaClj O.oi %, MgClz O.oi %,
NajHPOi 0.005 % and NaHCOj O.i %. Through the solution was
bubbled a mixture of 95 % Oj and 5 % COa.

In the following list of investigated drugs the number of micro -
grams of the respective drugs per c.c. fluid in the bath (y/ml), that
was as a rale found to be the most suitable dosage for the experiment,
is given in brackets.

Acetylcholine hydrochloride (O.oi), acetyl-/3-methylcholine hydro-
chloride= mecholyl (O.oi), acetyl-^-methylcholine bromide= mecholyl-
bromide^ (O.oi), carbaminoylcholine hydrochloride = doryl (O.oi),
pilocarpine hydrochloride (0.3), physostigmine (eserine) salicylate
(0.03), prostigmine methylsulphate (O.i).

1-epinephrine hydrochloride (O.oi), 1-corbasil hydrochloride (0.03),
d-corbasil base (0.9),, dl-axterenol bydrocbloride (0.04) dl-adrenalone
hydrocbloride (3), dl-epbedrine bydrocbloride (9), l-adrianol hydro-
chloride (0.6), 1-sympatol hydrochloride (0.3), dl-benzedrine sulphate
(l.r), dl-ventol sulphate (2.9).

272

HlKAlf RTDIN.

Atropine sulphate (O.oe), ergotamine tartrate (O.i), substance P,
Euler and Gaddum (1 — 6 E.), nicotine sulphate (2.8), histamine
(0.9), barium hydrochloride (3).

The changes in temperature were effected in the folio-wing maimer.
The result of such a dosage of the respective drug as gave a moderate
effect at a temperature of 38° C, -was first observed. Afterwards the
effect of the same dosage at a somewhat lower or higher temperature
(some of the temperatures ranging between 38° and 18° and, respec-
tively, between 38° and 40° in the bath fluid). The experiment was
then repeated at the original temperature of 38°. This procedure was
repeated the largest possible number of times (e. g. 38° — ^24° — 38° —
24° — 38°) on each piece of gut. At each temperature at least two effects
of the drug in question were studied before the temperature was changed.
After each change of temperature 10 minutes as a rule were allowed to
elapse before a drug was added to the bath.

The preliminary investigations showed that temperature variations
between 38° and the lower degrees of temperature stated above were
the most suitable for our approach to the problem. In the sequel,
therefore, it was mainly the excitability at these lower temperatures
which was compared with that at 38°.

The material used for the experiments consisted of pieces of gut
from rabbits. The results are uniform, apart from the cases where
the contrary is reported. In judging the residts, changes in the intestinal
tone and in the magnitude and rate of the movements as well as the
duration of the reaction have hitherto been taken into account, but
not such factors as the bearing on them of the concentration of the
poison and the latency period from the adding of the drug to the bath
fluid until the beginning of the response of the intestine.

Results.

The rabbit intestine was found to be rather insensitive to the
lowering of the temperature, but more sensitive to its rise.

The lowering of the temperature from 38° to 84° as a rule
entailed no observable change in the working of the intestine
during periods of observation of 10 — 15 minutes. At tempera-
tures of 32° — 30° — 28° — 26°, varying for different guts, the
contraction frequency in particular diminishes, but often, though
in a lesser degree, also the amplitude. Thus, whereas at 38° the
frequency is about 7 — 8 contractions per minute, the rate at
30° — 28°, as a rule, is about 3.4 per minute (Eig. !)• At tempera-
tures of 24° — 22° the intestine is still working, but usually with
a markedly diminished magnitude and rate of contraction, (Eig.
2). At 30° — 28° the magnitude of contraction is often 20 — 40

273

ACTION OF DRUGS ON AUTONOMIC EFFECTOR CELUS.

per cent, less than at 38°, sometimes, however, actually greater
than at the last-mentioned temperature. At 20°— 18° the responses
are minimal and irregular, or else the intestine is quite motion-
less (Fig. 2). As a rule, no change in the tonus could be observed.

The raising of the temperature in the water bath from 38°
to 40° — 41° as a rule entailed no observable change in the working
of the intestine during observation periods of 10 — 15 minutes.
If the temperature is further raised the contraction rate as a rule
increases (from about 8 to 10—12 contractions per minute). At

Fig. 1. Isolated rabbits intestine in Tyrode solution. Addition of 0.01 yjva.
aoetyloholine at 1) 38°, 2) 28° and 3) 38° C.

temperatures between 41° and 46°, varying for different guts
the amplitude falls to a minimum tvith a markedly slowed-down
and also often irregular frequency. The range of temperature
within which the working of the intestine changes from normal
until it completely ceases to function, is as a rule comparatively
small, being about 2° — 4°. Moreover, the intestine is often so
injured by the rises of temperature that repeated tests on the
same piece of gut cannot be made. As lowering of the temperature
was found to entail more favourable experimental conditions
for our approach to the problem, we have hitherto made merely
a few preliminary experiments with temperatures over 38°.

The addition of different drugs at varying temperatures to the
bath fluid gave the following results.

Acethylcholine: On the lowering of the temperature to 32° or
30 in the bath, intensified acetylcholine effects were often ob-
tained. On further lowering to 28°— 26° this intensification became
considerable, being as a rule of the magnitude and character
indicated in Fig. 1, thus with an appreciable increase of the tone.

18 — i60215. Acta phys. Scandinav, Vol.ll.

274

hAkan rtdtn.

In cases "witli less intensification of .tlie effect, it is manifested,
in some instances, by greater magnitude of contraction relatively
to tbe increase on an addition of acetylcholine at 38°, in other
cases merely by increased tone. The contraction frequency after
the addition of acetylcholine remained as a rule xmchanged, in
experiments both at 38° and at lower temperatures. In isolated
cases, however, a slight increase of the frequency after the addi-
tion of acetylcholine was observed, when it had previously been
markedly low owing to the reduced temperature in the bath.

Even at very low temperatures
(22° — 18°), in cases where the
working of the gut had actually
ceased, a marked contraction was
obtained on the addition of acetyl-
choline in a dosage which at 38°
had given a moderate response.

When the temperature was raised
above 38°, a moderate reduction
of the choline effect ensued at 40°
— 41°; at 42° — 44° this reduction
in the effect was pronounced and
general, as compared with the
effect at 38°. In those cases where
the working of the intestine at these high temperatures was
insignificant or nil, the action of acetylcholine doses which at 38°
had given a marked effect was completely neutralized. Kegularly
recurring intensification of the acetylcholine effect at any tem-
perature above 38° has not been observed.

A similar intensification of the intestinal effect at low tem-
peratures as after the addition of acetylcholine was also observed
in regard to pilocarpine (fig. 2) as well as doryl and mecholyl.
In many cases, however, the intensification was not so marked
as for acetylcholine. As regards pJiysostigmine (eserine) and
prostigmine, on the other hand, the said intensification of the
effect could not be observed at the lower temperatures; the res-
ponse of the intestine in many cases began to decrease even at
30° — 28° and then diminished according as the temperature was
further lowered.

Epinephrine showed an intensified effect at 30° as compared
with 38°, and in many cases it further increased at about 28°,
in order afterwards to fall again. At 22° — 24° the action of the

Fig. 2. Isolated rabbits intestine
in Tyrode solution. Addition of 0.3
y/ml pilocarpine at 1) 38°, 2) 24”
and 3) 38” C.

ACTION OF DRUGS ON AUTONOMIC EFFECTOR CELLS.

275

(

Fig. 3. Isolftted rabbits intestine in Tyrodo solution. Addition of 0.01 /'/ml
epinephrine at 1) 38®, 2) 24® and 3) 38®’ C.

same amount of epinephrine as had shoivn a marked affect at
38° was greatly •weakened or neutralized. "When the temperature
was raised, an increasing inhibition of the epinephrine effects
ensued, as compared with the action of that drug at 38°, beginning
as a rule at 41° — 42°. At 44° the effect of the epineplirine as a rule
was completely neutralized. The above-mentioned intensified
action of the epinephrine was manifested especially by dimin-
ishing magnitude of the contractions, reduced tonus and pro-
longed effect, in certain cases also by decrease of the rate (Fig. 3).

A similar intensified reaction at temperatures of about 28°
was sho'wn also by d- and l-corbasil, dl-arterenol, l-adrianol and
dUadrenalon. These intensifications, however, as a rule were not
so marked as for epinephrine. 1-sympatol, generally speaking, did
not produce this intensified effect: on two pieces of gut, how-
ever, sympatol repeatedly produced the strongest reaction at
28°, as compared -with its effect at 38°.

Ephedrine, benzedrine and veritol induced no intensified response
at temperatures below 38°. At temperatures of 30° — 26°, varying
for different guts, the excitability considerably decreased.

As in the case of ephedrine, the maximum reaction of the
intestine was at 38° for atropine, ergotamine, nicotine, substance P
and histamine (fig. 4).

Fig. 4. Isolated rabbits intestine in Tyrode solution. Addition of 1.2 y/ml hist-
amine at 1) 38°, 2) 28® and 3) 38° C.

276

HAKAN RYDIN.

Barium chloride. No certain cliange in the working of the in-
testine on the addition of barium chloride at 40° and 42° or 32°,
30° and 28°, as compared with its effect at 38°, could, as a rule,
be observed. In certain cases, however, a somewhat weaker ac-
tion was observed at those temperatures as compared with the
effect at 38°. The response of the intestine to barium chloride,
generally speaking, diminished rapidly at higher or lower temper-
atures than those just mentioned.

Discussion.

The results of this investigation show that the rabbit intestine
at different temperatures varies its response to the studied drugs
in different ways. The following survey seems to indicate that
Dale’s theory of neorohumoral transmission can usefully serve
as a working hypothesis in the interpretation of the results
hitherto obtained.

Acetylcholine and certain farasympathomiineiica in the proper
sense of the term, closely related to it in their mechanism of ac-
tion as well as epinephrine and certain closely related sympatho-
mimetica in contradistinction from the other studied drugs, thus
induce a more intense activity at lower temperatures (about 28°)
than normal (38°). The group of parasympathomimetica, however,
are distinguished from the sympathomimetica by the fact that the
intensified response remains even at such low temperatures
(24° — 20°) that the movement of the gut are irregular or have
completely ceased.

Among the parasympathomimetica, pilocarpine as well as
doryl, mecholyl chloride and mecholyl bromide, which are chem-
ically related to acetylcholine, likewise induced the said intensi-
fied effect at the lower temperatures, though scarcely in the same
degree , as the acetylcholine. In the case of physostigmine and
prostigmine this intensification failed to manifest itself. These
results conduce to bear out the view that the three first-mentioned
drugs (pilocarpine, doryl and mecholyl) act directly on the effector
cells in the same or similar way as acetylcholine, in contradistinc-
tion from physostigmine and prostigmine the inhibitors of choline-
esterase.

Among the sympathomimetica studied, corbasil, arterenol,
adrenalon and adrianol showed the same effect as epinephrine,
though not always so marked, whereas the so-called pseudo-

ACTION OF DRUGS ON AUTONOMIC EFFECTOR CELLS. 271

sympatliomimetica, ephedrine, benzedrine, vcritol, and, as a
rule, sympatol did not produce this intensification. A parallelism
between chemical structure and pharmacodynamic effect can be
noted here, in that the intensified effect at the lower temperatures
was obtained only as regards sympathomimetica containing a
hydroxyl group in the meta-position. The connection between
the structure and the action of the sympathomimetica has been
previously subjected to numerous investigations. Here it need
merely be mOntioned that Emilsson (1942), in investigations
regarding the inhibiting action on the isolated small intestine of
the rabbit found the same grouping of these B 3 rmpathomimetica
as has just been indicated.

Atropine, unlike acetylcholine, does not induce any intensifi-
cation of the effect on the intestine at about 28°. This observation
seems to bear out the view that atropine, in its prevention of the
muscarinic effects of acetylcholine and its esters does not directly
act on the same structures of the effector cells as acetylcholine.

The experiments with ergotamine seem likewise to indicate,
that this drug does not attack the same structure of the effector
cell as epinephrine.

Barium chloride, which stimulates muscles of all types, regard-
less of innervation, did not produce any intensified effect at the
low temperatures in question, nor did histamine or substance P.

In attempting to explain why the rabbit intestine at var 3 dng
temperatures changes its response to different drugs in diverse
ways, it seems plausible to suppose that different kinds of ‘'re-
ceptor substances^’, or reactive material of the effector cells, slma
dissimilar resistance to loio temperatures.

On the other hand, it seems less probable that the cause of the
intensified effect at the lower temperatures is to be sought in
diminished enzymatic breaking-down of these drugs (cholinester-
asis, amino-oxidasis etc.). The result of the eserin experiment
argues against such a supposition.

Preliminary tests with the same method on other organs (the
uterus of guinea pigs and rabbits, guinea pig intestines) have
not verified the results obtained on rabbits. The effects are also
small relatively to those on rabbits, nor can any certain difference
in the action of different groups of drugs be shown in regard to
these organs. It therefore seems as if, among the organs hitherto
tested, only the rabbit intestine were suited for studies of this
nature.

278

hAkan kydin

Summary.

Th.e action of a number of drugs on the isolated, surviving small
intestine of the rabbit was studied at different temperatures
according to the method of Magnus.

Acethylcholine and some closely related parasympathomimetica
(pilocarpine, doryl, mecholyl) showed at temperatures roundabout
28® a stronger effect on the intestine than at 38°. The intensified
action of those drugs was manifested even at such low tempera-
tures (24° — 20°) that the intestine worked irregularly or stopped
working altogether. On the addition of physostigmine and pro-
stigmine, on the other hand, these intensified effects were not
produced.

Epinephrine and certain sympathomimetica closely related to
it in structure (corbasil, arterenol, adrianol and adfenalon)
showed a maximum effect at temperatures roundabout 28°.
This intensified effect, however, was not retained at the lower
temperatures. As regards the so-called pseudosympathomimetica
(ephedrine, benzedrine and veritol) no increased effect was pro-
duced on lowering the temperature below 38°. The intensified
effect at the lower temperatures was obtained only as regards
sympathomimetica containing a hydroxyl group in the meta-
position.

As for atropine, ergotamine, histamine and substance P, no
intensified effect was induced.

The dissimilar response of the rabbit intestine to different
groups of drugs at varying temperature is presumably due to
the apparently dissimilar resistance of different kinds of
“receptor substances”, or reactive material of the effector cells,
to low temperatures.

Beferences.

Blaschko, H., and H. Schlossmann, J. Physiol. 1938, 92. 26 P.
Emilsson, B., Acta Physiol. Scand. 1941 — 1942. 3. 275. 335.
Emmelin, N., G. Eahlson and F. Wicksell, Ibidem 1941. 2. 123.
LAxmENDER, W., and B. Mertz, Arch. exp. Path. Pharmak. 1940.
194. 389.

Eentzhog, U., Personal communication.

Zadina, B., C. R. Soc. Biol. Paris 1939. 132. 28.

From the Pharmucologicol Department, Karolinskn Institutet,

Stockholm.

On tlie Synthesis of Proteins in Rat hy Dialyzed
Casein Digests.

K A. J. WRETLIND.

Eeccivcd 21 January 1946.

It h.as been shown among others by Henriques and Hansen
(1905) that nitrogen equilibrium and thus also protein synthesis
are easily maintained in rats by enzymatic hydrolysed casein
as the sole rdtrogen-containing food. In order to make such casein
digests free from imdigested proteins and high-molecular pep-
tides, the author (1944, 1945) has described a method for dia-
lyzing the hydrolysates through cellophane membranes. By this
procedure a product is obtained which contains 80 — 85 %free amino
acids and 15 — 20 % low-molecular peptides. It was in order to show
that amino acids produced in this manner are able to maintain
nitrogen equilibrium and also to enable protein synthesis in rats
that this investigation was made.

method.

For the experiments adult rats have been used. The animals
were kept in cages of the kind described by Henriques and
Hansen (1. c). To the vessel used for collecting urine from the
rat was added 10 ml of a saturated solution of boric acid. Every
day at ten o’clock the urine was collected and diluted to 200 ml
with distilled water. From this solution 20 ml was taken for
nitrogen determination. The faeces were transferred directly to

280

K. A. J. ^rKETIJ:^D.

the combustion vessel. The nitrogen of both urine and faeces
•was determined according to the method of Kjeldahl as described
by Pbtebs and van Slyke (1932).

As nitrogen-free food, butter -was used together -with sugar and
salts. The proportions of the ingredients are sho-wn by the protocol
of each experiment. The butter and sugar (cane-sugar) -were
ordinary commercial products. As salt mixture, the one described
by OsBOENE and Mendel (1919) was used. The butter was melted
and the sugar and salts were added while stiiTing, which was
continued until the butter had solidified. It was found that this
composition contained small amounts of nitrogen, the amount
of which was determined in each experiment.

In order to obtain solid faeces the rat daily got cellulose in the
form of filter-papers cut into small pieces and mixed with the
other food. By analysis it was found that the cellulose contained
0.33 mg nitrogen in 1 g. Also in this case the nitrogen was de-
termined according to the method of Kjeldahl.

An amino acid preparation made by the above-mentioned
method was used.^ The nitrogen content of this preparation was
12.7 %. The amino acids were made to a 26 % suspension by
adding distilled water. The pH of the suspension was about 7.1.
The nitrogen-content of this suspension was estimated by the
method of Kjeldahl. To the food was added amount corresponding
to the nitrogen which may be found in the tables.

The amount of the butter, sugar and salt mixture per diem
must be large enough to ensure the needed calories but not so
large that the rat does not eat all its food in 24 hours. This is a
rather difficult task, but •with some experience one may find the
amounts needed for rats of different weights.

Cellulose was given in amounts of 0.5 to 1.0 g. Ko difference
could be observed in the faeces whether 0.5 or 1.0 g was
used.

As a rule, the rats got the amino acids in quantities corres-
ponding to from 130 to 182 mg nitrogen, which is a good deal
more than the minimal output of nitrogen. This was done to
show the nitrogen retention more clearly.

Water was given ad libitum.

* The commercial product, Aminosol, made by this method has been used in
this investigation. Aminosol is prepared by Titrum, Stockholm.

SYNTHESIS OF PROTEINS IN RAT.

281

Experiment I.

An adult rat weigliing 115 g was taken directly to the experi-
ment from an ordinary diet. As nitrogen-free food, a mixture of
250 g of butter, 40 g of sugar and 15 g of salts was used. The rat
was given amounts of this composition ranging between 4.00
and 6.00 daily. 1 g of this mixture contained 0.893 mg N. 1 g cel-
lulose per diem rvas also given corresponding to 0.33 mg N. The
amounts of nitrogen given as amino acids will be found in the
table.

Table I.

Date

194.’)

"Weight of
tho rat

cs ^

gt;
s 1 «

^ "3

»<•

C u
•»-> a: 0

^ 0 tT ”

mg N in the
aminoadflo

mg N total
given

.2.5

75 c

mg N total
excreted

mg N
Difference
between N
given and
excreted

5. 7.

115

6.00

5.09

133.0

1.39.29

49.84

39.81

89.16

+ 50.14

6. 7.

117

6.00

5.09

133.0

139.29

71.40

33.49

104.89

-f 34.40

7. 7.

119

4.00

3.90

80.19

84.09

73.04

29.12

102.70

- 18.67

8. 7.

111

5.00

4.80

106.9

111.70

79.80

50.90

130.70

- 19.00

9. 7.

111

6.00

5.09

133.0

139.29

72.80

23.80

96.10

+ 43.19

10. 7.

6.00

5.69

133.0

139.29

57.60

29.12

86.02

52.07

11. 7.

6.00

5.09

133.6

139.29

77.06

49.00

126.58

-j~ 12*71

12. 7.

113

6.00

5.09

133.0

139.29

65.20

42.22

107.48

-t- 31.81

13. 7.

113

6.00

5.09

133.0

139.29

66.82

49.60

116.82

-f 22.97

14. 7.

115

6.00

5.09

133.0

139.29

87.09

39.81

126.40

+ 12.89

15. 7.

117

6.00

5.09

133.0

139.29

97.02

30.77

127.79

-P 11.60

16. 7.

118

6.00

i).C9

133.6

139.29

77.70

43.08

121.88

+ 17.91

17. 7.

120

6.00

5.09

106.9

112.09

66.82

29.12

95.94

-f 16.06

From table I it may be seen that during the 13 days of the
experiment the rat retained 269.11 mg nitrogen, corresponding
to 1.7 g of proteins. As the rat gained 5 g in weight it may be
assumed that it got enough calories.

Experiment 11.

The rat weighed 140 g. The butter, sugar and salt mixture was
made up of 240 g of butter, 80 g of sugar and 15 g of salts. 1 g of
this mixture contained 0.812 mg N, As in experiment I, the rat
got 1 g of cellulose (0.33 mg N) daily. The amino acids were given
as in experiment I. However the rat did not get any amino acids
on the first day of the experiment. In table II it is found that in

282

K. A. J. WRETLIND.

14: days (from 20.10 to 2.11) the rat retained 467.66 mgN, corres-
ponding to 2.9 g of proteins. In this experiment the rat gained
7 g in weight.

Table n.

SYNTHESIS OF PROTEINS IN RAT.

283

and II. The composition of the butter, sugar and salt mixture
was 240 g of butter, 80 g of sugar and 15 g of salts. Of this mixture
amounts of 5.0 g were given. 1 g of this compound contained
0.812 mg N. Besides this 0.5 g of cellulose, containing 0.17 mg
N was given. This experiment was broken off after 9 days be-
cause the rat did not eat its food.

From table III it may be seen that the small amounts of nitrogen
in the butter sugar and salt composition and in the cellulose are
of no importance with regard to the nitrogen equilibrium and the
nitrogen-retaining power of the amino acids in the experiment
I and II. The nitrogen quantity given in the butter, sugar and
salt mixture and the cellulose was always below 10 % of the
excreted amount of nitrogen, and it is thus impossible to get a
nitrogen equilibrium with this small amount of nitrogen.

Summary.

It has been shown that dialyzed enzymic casein hydrolysate
could give not only nitrogen equilibrium but also a considerable
nitrogen retention in rats.

Literature,

Henriques, V., and C. Hansen, Hoppe-Seyl. Z. 1905. 43. 417.
Osborne, T. B., and L. B. Mendel, J. Biol. Chem. 1919. 37. 572.
Peters, J. P., and D. D. van Slyke, Quantitative Clinical Chemistry.

Vol. II. London 1932. page 516.

Wretlind, K. a. j., Svensk Lakartidn. 1944. 41. 1033.

— , Nord. Med. 1944. 23. 1762.

— , Ibidem 1945. 27. 1827.

From the Department of Physiology, University of Lund.

Effect of Acetylcholine and Adenosine Triphos-
phate on Denervated Muscle.

FRITZ BUCHTHAL and &EORG KAHLSON.

Received 2 February 1946.

In previous papers (Bvcrtbal et al. 1944 a b c) it could be
demonstrated that minute amounts of adenosine triphosphate
(ATP) release contraction in frog and mammalian striated muscle.
It tvas suggested that ATP constitutes a further link in the chain
of reactions initiated by the injection of acetylcholine into the
artery of a muscle. This high energetic phosphate is even effec-
tive after complete curarization. Apart from the release of con-
traction by ATP another property of this substance was dis-
closed, viz. its sensitizing effect on subsequent applications of
acetylcholine. Thus, it seemed likely that in some way or other
more intimate interactions take place in the nerve muscle system
between acetylcholine and ATP. We thought it of interest to
investigate the effect of ATP on denervated muscle, as denerva-
tion causes profound changes in the reaction to acetylcholine.
It is generally accepted that normal skeletal .muscle is relatively
insensitive to intra-arterially injected acetylcholine, while de-
nervated muscle after application of minute amounts of this
substance exhibits a considerable increase in tension and dura-
tion of contraction.

Method.

The experiments were performed on the anterior tibial muscles of
the cat under chloralose or decerebrate as described by Beowjt (1938).
The substances were applied by close arterial injection into the distal
part of the anterior tibial artery, the proximal part being temporarily
closed by traction on a ligature, when the substances are apphed. The
tension developed by tibial muscles on intra-arterial injection of ace-
tylcholine may show rather large individual differences in different
cats. It therefore seemed of special interest to record alternately within

EFFECT OF ACETFLCHOLINE AND ADENOSINE TRIPHOSPHATE. 285

Bliort intervals the tension developed by normal and denervated muscles,
thus enabling a quantitative comparison under standardized conditions.
Therefore, both tibial muscles were mounted in the Brown-Schuster
myograph exerting tension on the same isometric lever.

Section of the right sciatic nerve was performed 6 — 85 days before
the experiments in all 25 cats. The completeness of the denervation
was controlled by electrical stimulation of the peroneal nerve. The
reduction in weight of the denervated muscle amounted to 25 — 60
per cent as compared with the normal muscle, depending on the time
allowed for degeneration.

All substances were applied iso-osmotically by substituting an equal
amount of NaCl + water in the Tyrode solution by the staple solution
of ATP. The pH of the injected solution was 7.3 and its temperature
37° C. The ATP was applied as sodium salt (for description of prepara-
tion and analysis cf. Buchthal et al. 1944 a). The substance was kindly
promded by Dr. A. Deutsch, research laboratory A. B. Leo, Hiilsing-
borg.

Eesults.

1. Sensitivity to acetylcholine.

Close arterial injection of minute amounts of acetylcholine into
denervated muscle releases a double mechanical response, con-
sisting of a quick initial phase followed by a protracted develop-
ment of tension. The quick phase is generally absent at the second
injection, but may exceptionally persist during the first three
applications. This is in agreement with the observations of Brow
( 1937). From the 6th day onwards after denervation no system-
atic correlation exists between duration of tension development
and time allowed for degeneration. However, in 3 of 26 experi-
ments with a degeneration period of 46, 47 and 52 days respec-
tively, an exceptional behaviour was observed, the muscle res-
ponding to acetylcholine solely by contractions of normal type
and duration. In another series of experiments comprising 3
animals with a similar degeneration time of 46, 47 and 49 days
resp. the denervated muscle responded with contractions of the
protracted type.

The present technique allows a direct comparison of the sen-
sitivity of the normal and denervated muscle to acetylcholine on
one and the same animal. The difference in threshold was con-
siderable, the denervated muscle being 20 — 200 times more
sensitive than the normal.

The blocking action of acetylcholine on subsequent injection
of this substance is well known from normal niuscle. In de-
nervated muscle it is even present when injections are made with
intervals of several minutes, denervated muscle becoming in-

286

FEITZ BUCHTHAL AND GEORG KAHLSON.

-1/'-

S.s^ 0 O!-^ S.S^ O.Oly Roly R Oly O.Oly O.Oly 3Sy O.Oly iSy 0 Oly 3 iy O.Oly

rffiTiri'nfiiiiiiii'lli‘J'iii'/iii'lilvi(ial’ilJtlii!i.!il^J'Jii*iil'>ui'«i'iiii'‘i<'>^‘l''>'''‘i''' I’-*' liiiiHilinVinnnWiii

„ — II 1/ — u V ■- j : — V 1 'J V — i

Fig. 1. Cat under chloralose. Mechanical responses from right (r) and left (1)
anterior tibial muscles to close arterial injection of acetylcholine. Right muscle
denervated four weeks previously. Time marks 1 per second. To the right: direct
electrical stimulation of the denervated muscle. Weights: 1 = 6.4 g; r = 3.0 g.

sensitive to acetylctoline after 10 — 20 injections. The response
to direct electric stimulation is retained (Fig. 1).

2. Effect of Na-adenosine trifhospliate.

A solution of ATP injected into the artery in amounts causing
submaximal responses evokes contractions of different types in
normal and denervated muscles. Compared with the response in
normal muscle ATP contractions of denervated muscles are con-
siderably prolonged, the difference in duration being of a similar
ratio as that observed with acetylcholine (Fig. 2). As is the case
for the latter, duration of contractions released by ATP, from
the 6th day onwards, is independent of the time allowed for
degeneration. In ATP contractions there is no quick phase even
with the first injection. The peak of tension developed during
contractions, with the same dose, was generally 30 — 50 per cent
higher in denervated — though atrophied — muscle than in the
corresponding normal miiscle. A definite small amount of ATP,
inactive in the normal, causes a pronounced contraction in the
corresponding denervated muscle (Fig. 2). There is, however, no
parallelism in the sensitivity of normal and denervated muscle
to acetylcholine and ATP, Thus, it was frequently observed that
a muscle relatively insensitive to acetylcholine developed strong
tension when ATP was injected or vice versa.

In denervated muscle, contrary to findings in normal, previous
injection of ATP does not enhance the tension developed by
subsequent application of acetylcholine.

EFFECT OF ACKTVLCUOMKE AND ADENOSINE IlUl'nOSPlIATE. 287

In the course of this investigation evidence nccumulnted thnt
under certain conditions ATP iras ineffective in eliciting mechanical
responses in denervated mvsde. This was always the case when
previously to the application of ATP acetylcholine had been
injected into the muscle.

The response to acetyl-
choline and to direct
electrical stimulation is,
however, retained. For
one experiment of this
t}’pe a .supply of lithium
ATP was available and it
was striking that Li -ATP
wa.s active even after pre-
vious injection of acetyl-
choline when the muscle
was refracK)r 3 ’ to Na-ATP.

This is in agreement with our observations on smooth muscle where
Li-ATP proved effective in cases whore the action of Isa-ATP was
strongly reduced (BucaTHAD and KAin.soN 1914).

Discussion,

The aim of thi.s investigation wa.s a direct comparison of the
reaction of normal and denervated muscle to injections of ace-
tylcholine and ATP. Independent of the time of dogenorntion
both substances release a contraction of long duration in de-
nervated muscle. It is well known thnt acetylcholine, ineffective
in initiating contractions in curarized miKscIe, evokes a .special
t37pe of response in denervated muscles. Since curarization, in
normal muscle, does not alter the tj'pe of contraction, it is obvious
that denervation, apart from changes in tlio peripheral nerves
and motor end plates causes changes in the reaction of the muscle
substance itself to chemical stimulation. It is on the other hand
worth mentioning that the greater sensitivity of denervated
muscle to chemical stimuli is abolished b 5 '^ curarinc (Buown 1937)
and even after a degeneration period of almost 3 months, the
motor end plate still is liigldy sensitive to acetylcholine.

The depressing effect of acetylcholine on subsequent applica-
tion of ATP in denervated muscle indicates that acetylcholine
interferes with the reaction of the muscle substance. As the re-
sponse to acetylcholine is retained when the preparation has be-

O.S'mm /m tnH) O.Sf S.Sy

ftrP/m />rP au,.

rnhoiluasmiimmmmm.: r: '.ryr. — -r -- — ,

Kii;. 2. Cnt nntJpr chlomloso, Jfechnnicnl rc-
uponsfs from riglit (r) nnU left (1) nntcrior tiliinl
muscIcH to close nrterinl iiiit'clion of ncotylclioline
nnd ATI’, llight muscle (lenerciitcsi five weeks
previously. Time marks 1 every ten seconds.
IVeighls; 1 .-i 7.0 g; r ^ 2,8 g.

288

FRITZ BDCHTHAL AND GEORG KAHLSON.

come refractory to ATP, we are forced to suppose that in denerv-
ated muscle acetylcholine prevents the interaction of the intra-
arterially applied ATP with the contractile substance. The-mechan-
ism of this inhibition remains obscure. It may, however, be noted
in this connection that acetylcholine actually interferes with the
enzymatic activity of myosin. Acetylcholine has a considerable
inhibitory influence on the adenosinetriphosphatase. Pufthermore,
previous application of acetylcholine abolishes the changes in
birefringence produced by ATP in normal frog muscle fibres.
This, too, indicates that acetylcholine, apart from its effect on
the motor end plate, in some way reacts on the contractile protein.

Summary.

Direct comparison of denervated with normal anterior tibial
muscles of the cat showed;

1. Sodium adenosine triphosphate initiates contractions in de-
nervated muscle which last considerably longer than in normal
muscle.

2. Previous application of acetylchoUne to denervated muscle
abolishes its sensitivity to ATP, while the reaction of normal
muscle to this substance is uninfluenced by previous injection
of acetylcholine.

3. In agreement with previous investigations we find denerv-
ated muscle considerably more sensitive to acetylcholine (average
40 — 50 times) than normal muscle; the curare-lihe action observed
on repeated applications of acetylcholine is highly pronounced
in denervated muscle. Since the increased sensitivity still is present
when degeneration has been allowed to proceed for 75 — 85 days,
and since this effect can be abolished by curarization, it must be
concluded that the motor end plate or part of it survives even
at this high degree of muscle atrophy.

References.

Brown, G. L., J. Physiol. 1937. 89. 438.

Brown, G. L., Ibidem 1938. 92. 22 P.

Buchthal, F., a. Dedtsch, and G. G. Knappeis, Acta Physiol. Scand.

1944 a. 8. 271.

Buchthal, F., and B. Folkow, Ibidem 1944 b. 8. 312.

Buchthal, F., and G. Kahlson, Ibidem 1944 c. 8. 317, 325.

From the Institute of Medical Chemistry, TJnivcrsily of TJpsala,

Sweden.

A Modified Preparation of tlie Universal Buffer
Descritoed by Teorell .and Stenlmgen.

SVEN bSTLING and PEKKA VIRTAMA.

Received 4 February 1946.

The universal buffer solution for the pH-range 2 — 12 described
by Teorkll and Stenhagen 1938 has been used in various
investigations in this laboratory and has apparently also been
found useful in other laboratories. The buffer is intended prim-
arily for biological and surface chemistry work. It is nitrogen-
free, the content of cations other than hydrogen ions is constant
throughout its range and the variation of buffer capacity and
ionic strength fairly moderate. The stock solution for the buffer
contains sodium phosphate, citrate and borate, together with an
excess of sodium hydroxide. In order to prepare a buffer solu-
tion with a desired hydrogen ion concentration, a certain volume
of stock solution is taken, the appropriate amount of hydro-
chloric acid added and the mixtiu-e diluted to a standard volume.
As originally described, the sodium phosphate and citrate needed
for the stock solution are obtained by titrating 100 ml of l-N
sodium hydroxide with a solution of phosphoric and citric
acid respectively. This procedure has to be used when suitable
buffer salts are not available. When buffer salts according to
Sorensen are accessible it is more convenient to use the latter
and thus avoid the titration procedure. We have therefore pre-
pared a series of stock solutions in this manner and, as the two
modes of preparation may not be exactly equivalent, remeasured
the buffer over the entire pH-range.

19 — i60215. Acta phys. Scandinav. Vol.ll.

290

SVEN OSTLING AND PBKKA VIKTAMA.

Experimental.

Polentiometric measurements. The measurements were carried
out at a temperature of 20.0° ±0.1° (’well-stirred water bath)
with the use of cells consisting of hydrogen electrodes (palladium
black electrolytically deposited on platinum foil, cf. Clark
1928) and 3.5-N calomel electrodes. A series of 3.5-N calomel
electrodes were prepared following the directions given by Bjer-
EUM and UNJLA.CK 1929. Calomel was prepared electrolytically
from purified mercury. The calomel electrodes did not differ
by more than 0.1 millivolt from each other and they had a po-
tential of 0.3717 volt against the hydrogen electrode in a solution
of 0.01-N HCI, 0.09-N NaCl (Veibel’s solution) in exact agreement
with that found by Bjerrum and Unmack. The hydrogen for
the bubbling electrodes w'as taken from a commercial cylinder
and purified by passing through an electric oven with a filling
of copper to remove oxygen. Before entering the electrode vessels
the hydrogen was saturated with moisture by passing through
a washing bottle filled ■with conductivity water and immersed
in the bath. The potential measurements were carried out using
the “Slide wire potentiometer” and Weston standard cell made
by the Cambridge Instrument Co. and a Lange’s Multiflex gal-
vanometer of high sensitivity. Eeadings were taken to 0.1 milli-
volt. Eor calculation of pH- values, the potential of the S.S-hf
calomel half-cell was taken as 0.2522 volt. For a discussion
of the standardization of the pH-scale the reader is referred to
a paper of Mac Iknes, Belcher and Shedlovsky 1938. The
liquid junction potential of the 3.5-N KCl-bridge was neglected.

Preparation of Buffer Solutions.

0.1'N Hydrochloric acid. Constant boiling hydrochloric acid
prepared according to Hulett and Bonner 1909 is diluted to
0.1-N. If the distillation has been carried out at 760 mm pres-
sure, 18.019 g of constant boiling acid is diluted to 1,000 ml.
The amount to be taken, when the distillation has been performed
at other atmospheric pressure may be obtained from tables given
in the paper of Hulett and Bonner or van Slyke and Peters
1932.

1-N Sodium hydroxide. A saturated solution of sodium hydroxide
(Eka, analytical reagent) is prepared and centrifuged in glass

MODIFIED PREPARATION OF THE UNIVERSAL BUFFER. 291

tubes provided witb rubber stoppers until all insoluble carbonate
has been collected at the bottom of the tube. 58 ml of the clear
supernatant solution is diluted with carbon dioxide-free con-
ductivity water to 1,000 ml. The strength of the solution thus
prepared is determined by titration with the 0.1-N hydrochloric
acid described above. It is very important that the acid aud the
base correspond exactly to each other.

Buffer stock solution. In a 1,000 ml measuring flask are placed
8.903 g (0.05 mole) of disodium phosphate (!N’a 2 HP 04 , 2 HjO,
Kahlbaum, “nach Sorensen” or equivalent), 7.00 g (0.0333 mole)
of crystallized citric acid (C 8 H 8 O 7 , 1 H 2 O, Riedel-de Haen or
equivalent), 3.54 g (0.0507 mole) of crystalline boric acid (H3BO3,
Kahlbaum or equivalent), 243.0 ml of 1-N sodium hydi’oxide
and carbon dioxide-free distilled water (preferably conductivity
water from a special still, such as that described by Ellis and
Kiehl 1935) added to make the volume 1,000 ml. The stock
solution is kept in a flask of Pyrex of Jena Gerate glass provided
with a 20 ml automatic pipette. The solution must be protected
from atmospheric carbon dioxide by efficient soda Hme tubes.

Buffer solutions of desired pH. A buffer solution is prepared
as follows: to 20 ml of the stock solution is added the amount
of 0.1-N hydrochloric acid found from table I for the pH in
question, and carbon dioxide-free distilled water added to make
the volume 100 ml. For example, if a buffer solution of pH .760
is wanted, then 30.33 ml of 0.1-N hydrochloric acid is added to
20.00 ml of stock solution and water added to 100 ml.

Table I,

pH-^

•00

•10

•20

•40

•50

•60

•70

•90

72.10

69.25

66.87

64.90

63.25

61.77

60.48

59.29

58.29

57.49

56.76

56.05

55.42

54.83

54.28

53.72

53.17

52.61

52.07

51.52

51.00

50 46

49.92

49.40

48.88

48.35

47.81

47.28

46.72

46.18

45.64

45,10

44.54

4.3.99

43 40

42.77

42.15

41.51

40.89

40.28

39.66

39.02

38.81

37.64

36.73

36.02

35.36

34.72

34.13

33.51

32.97

32.46

31.90

31.36

30.82

30.33

29.88

29.45

29.06

28.70

28.44

28.20

27.91

27.56

27.20

26.83

26.34

25.77

25.12

24.48

23.82

23.21

22.60

21.95

21.32

20.71

20.13

19.60

19.10

18.65

18.24

17.84

17.51

17.20

16.92

16.68

16.35

15.98

15.66

15.09

14.59

13.92

13.08

12.09

10.76

292

SVEN OSTLING AND PEKKA VIRTAMA.

The modified preparation should give a buffer solution of
the same molar strength as that given by the original mode of
of preparation, i. e.

H,P04

H3O3

HaCi

NaOH

HCl

O.OlOOj

0.01141 whole range

0.0067 °

0.069

0.001 (pH 11.50) to 0.072 (pH 2.00)

Fig. 1. Variation of pH with addition of 0.1-N HCl to buffer stock solution.

The measurements show small systematic differences, however.
In the whole range the modified preparation gives solutions,
which are slightly less acidic than those of the original prepara-
tion (see fig. 1). Part of the difference (0,032 pH units) is due
to the new standardization of pH-scale, the rest is connected
with the titration procedure in the original mode of preparation
which is likely to give the stock solution a phosphoric acid content
slightly higher than that of the modified preparation. The pH-
values of the original buffer have been repeatly checked in the
Institute of Physiology and confirmed ■svithin the unsystematic
experimental error of 0.03 — 0.06 pH units. The measurements
on the modified preparation have been carried out on five dif-
ferent stock solutions, and the agreement between the different
series of measurements is within 0.05 pH units.

MODIFIED PREPARATION OF THE UNIVERSAL BUFFER.

293

The variation in ionic strength with pH is shown in fig. 2.
We are indebted to Professor T. Teorell for the loan of a
Multiflex galvanometer and to Docent E. Stenhagen for advice.

References.

Bjeurum, N., and A. Unmack, Elektrometrische Messungen mit
Wasserstoffelektronen in Mischungen von Sauren und Basen
mit Salzen, Copenhagen 1929. 24, 30 and 31.

Clark, W. M., The determination of hydrogen ions, 3rd Ed. London
and Baltimore 1928. 286.

Ellis, S. B., and S. J. Kiehl, J. Amer. Chem. Soc. 1935. 57. 2145.

Hulett, 6 . A., and W. D. Bonner, Ibidem. 1909. 31. 390.

hlAC Innes, D. a., D. Belcher, and T. Shedlovsky, Ibidem. 1938.
60. 1094.

VAN Slyke, D. D., and J. P. Peters, Quantitative Clinical Chemistry,
London and Baltimore 1932. 2. 27.

Teorell, T., and E. Stenhagen, Biochem. Z. 1938. 299. 416.

From XJniversitetets Biokemiske Institut, Copenhagen.

A Note on tlie Biogenesis of Choline and
Creatine.

GUNNAR STEENSHOLT.

Received 14 Pebmary 1946.

In recent years much, work has "been expended on the study of
the hjochemistry of choline, and our knowledge of this remark-
able compound is now fairly extensive. One of the most interesting
problems encountered in this field of research is that of the bio-
genesis of choline. The older literature on the problem has been
aptly summarised in the book by Guggenheim (1940), and the
more recent contributions have been discussed by Weele (1943)
(cf. also Steensholt 1945 a). A historical review of the subject
ivill therefore not be given here.

The pioneer investigations by uu Vigneaud, Chanelee,
Cohen and Beown (1940) which led to the recognition of the
essential role played by methionine in biological methylation
processes, were carried out on intact animals by the method of
isotopes. To the present writer it appeared not unimportant to
carry out investigations of the same reactions by means of tissue
slices or tissue extracts and more direct methods of choline de-
termination, and a report upon some results obtained in this
direction was recently given (Steensholt 1945 a). By in vitro
experiments it was found that muscle and liver tissue in the rat
are able to catalyse the methylation of ethanol amine to choline,
the methyl groups being furnished by methionine.

In this work choline was determined by the method of IMaeenzi
and Caedini (1942). The principle of this method is to precipitate
choline as reineckate, and then determine the chromium in the
reineckate by a procedure based on the reaction of Cazeneuve
(1900).

BIOGENESIS OF CHOLINE AND CKEATINE.

295

Of the other methods of determination of choline we mention
here that of Beattie (1936) and of Koman (1930). Beattie’s
method consists in precipitating choline as reinecke salt and
measuring colorimetrically the red colour imparted by this salt
to acetone. The method of Roman is based on previous work
by, Stanek, who found that choline can be precipitated by
potassium triiodide. A choline periodide is formed, which was
recognised as a choline ennea-iodide. These findings were con-
firmed by Roman. In his method choline is precipitated in the
form mentioned, and the iodide determined by titration with
sodium thiosulphate.

The purpose of the present paper is to compare the results on
chohne synthesis- previously obtained by the method of Maeenzi
and Cardini to those obtained by the methods of Beattie and
Roman, and thus to corroborate and confirm our previous findings
(Steensholt 1945 a).

In all the work referred to above only methionine has been
used as methyl donator. However, according to investigations by
Toennies and Kolb (1939) methionine is easily oxidised to the
sulphoxide. This conversion is easily effected by means of hydrogen
peroxide, and it is far easier to oxidise methionine than cystein
and cystine [Toennies and Callan (1939)]. We have therefore
to consider the possibility of methionine sulphoxoide acting as a
methyl donator. In this connection it is very interesting that
Bennett (1941) has shown, by feeding methionine sulphoxide
to animals, that it is capable of completely replacing methionine.
On the basis of our present knowledge we are entitled to assume
that the biological action of methionine is due in large part to
its methylating properties, and obviously Bennett’s work leads
us to believe that the sulphoxide acts in the same way. The
objection might be raised that methionine sulphoxide is reduced
to the sulphide in the animal body. However, it is an experimental
fact that the sulphoxide is very resistent against reduction by
hydrogen iodide. It is therefore quite probable that the sulphoxide
exerts its action in the animal organism without previous reduction.

On considering this evidence it may not appear superfluous to
investigate by in vitro experiments the properties of methionine
sulphoxide as a methylating agent. A report on some experiments
m this direction will be presented in the present paper.

Methionine sulphoxide is also of interest for another type of
methylation processes, which has claimed’ the attention of bio-

296

GUNNAU STEENSHOLT.

ch.emists in recent years, namely the biosynthesis of creatine. It
is generally believed today that creatine is formed in the animal
body mainly by methylation of guanidine acetic acid, the methyl
groups being furnished by a suitable donator, probably methionine.
The present ■writer recently had occasion to give a short review
of these developments together "with some ne'w experimental
results on the biosynthesis of creatine [Steensholt (1945 b)].
From "what has been said above it is natural to raise the question
■whether also methionine sulphoxide is capable of supplying methyl
groups for the methylation of guanidine acetic acid. In the present
note a report ■will be given on some results recently obtained on
this problem.

Part of our previous ■work in this field ■was carried out by a modi-
fication of the "well kno-wn method of Folin for the determination
of creatine. Due to the. defects of this method, in particular
its comparative unspecificity, the work was later supplemented
by investigations based on a method of creatine determination
developed by Benedict and Bbhee (1936), Langley and Evans
(1936) and Lbhnartz (1941). For the convenience of the reader
and for the sake of accuracy in the description of experiments,
a brief summary of these methods ■will be given below.

Experimental Part.

Biological material. Rat liver and rat muscles were used throughout.
The animals were from 3 to 5 months old, and had been kept on a diet
believed to be sufficient in all respects. They were killed by decapita-
tion and the organs to be used removed immediately after death.

Substrates. The methionine was a Hoffmann-la Roche product. The
ethanol amine was partly a commercial product, carefully purified by
repeated distillations, and partly synthesised by the -writer according
to Knorb (1897). Briefly described this method consists in leading a
stream of ethylene oxide through a concentrated aqueous ammonia
solution and subsequently fractionating the reaction mixture. Both
samples were found to behave in exactly the same way in all the ex-
periments described below.

The methionine sulphoxide was synthesised according to Toennies
and Kolb (1939) by oxidising methionine by means of hydrogen per-
oxide in a suitable solvent. The method was found to work very satis-
factorily.

Further the guanidine acetic acid used in the present work was
s)Tithesised according to Nencki and Sieber (1878).

Determination of choline. The method of Mabenzi and Cardini is
as follows;

BIOGENESIS or CJIOLINE AND CREATINE.

297

The amount of choline to bo determined varies between 15 and 100 y.
The volume of the sample may range from 1 to 3 ml. The sample is
placed in a centrifuge tube with slender end and an equal volume of
an aqueous solution of ammonium reineckate is added. The mixture
is cooled in ice for 20 minutes, and the precipitate is then spun down
in the centrifuge. The supernatant liquid is removed, and the precipi-
tate is washed two or three times with 0.5 ml ice cold aclohol. It is
then dissolved in 1 ml acetone and the solution is transferred to an
ordinary test tube, the centrifuge tube being carefully washed with
some 60 % acetone. IVo then add: 2 ml of water, 0.2 ml 10 % NaOH
and 0.1 ml pcrhydrol for each 50 y choline in the sample. The tubes are
now placed in a boiling water bath for about half an hour. After cooling
2 ml of sulphuric acid are added together with sufficient diphenyl
carhazide (in a 0.2 % alcoholic solution) to give a final concentration
of 8 %. The mixture is finallj’ transferred to a measuring flask and
diluted to a suitable volume, in our ivork to 25 ml. The colorimetric
measurements were carried out with the Pulfrich photometer, using
filter S 53. The comparison tube contained a blank consisting of 2 ml
sulphuric acid and 2 ml diphenyl carbazidc solution made up to a final"
volume of 25 ml.

The method of Beattie was applied in the following form:

To 2 ml of the solution to be anal)'sed is added 2 ml of a saturated
solution of ammonium reineckate, and the precipitation is completed
by cooling in ice for about 20 minutes. The precipitate is separated by
centrifugation and washed with 2 ml of ice cold water and twice with
2 ml of absolute alcohol. It is then dissolved in a suitable amount of
aceton and the colorimetric measurement carried out in a photoelectric
photometer (or according to the instructions given by Beattie in
her paper).

The method of Roman was as follows:

The solution in which choline is to be determined must be neutral
or weakly acid. Hence the trichloroacetic acid extracts from our
biological experiments were neutralised with the appropriate amount
of NaOH. The rest of the determination was as follows:

1 ml of the solution to be anal 3 'sed was placed in a centrifuge tube
with slender end, and 0.3 ml of a precipitation reagent (made by dis-
solving 3 g of iodine in 100 ml 1-n iodine solution) were added. A brown-
ish precipitate forms, which is separated from the liquid by centrifuga-
tion for 10 minutes at 3000 R. P. M. The precipitate is washed 3 times
with ice cold water. Then 1 or 2 ml of chloroform is added, and the
titration is carried out Avith n/100 sodium thiosulphate solution. It
is a two phase titration, and thorough shaking during the titration is
necessary.

Determination of creatine. We have throughout determined the
amount of total creatinine in the reaction mixtures. Por this purpose
Ave have partly used the method of Polin in a modification due to
Lieb and Zachekl (1934 a & b):

2 ml of the solution to be analysed were pipetted into a 10 ml measur-
ing flask. 1 ml n-hydrochloric acid Avas added and the mixture heated

298

GUNNAR STEENSHOLT.

for 20 minutes in an autoclave at 130° C. After cooling 0.4 ml 10% NaOH
and 4 ml of a saturated picric acid solution were added, and the mixture
left standing for 10 minutes. Colorimetric measurements were then
carried out with the Pulfrich photometer using filter S 53.

The technical details of the other method used for determination
of creatine will be described below in connection with the report on a
typical experiment.

Synthesis of choline. In order to save space we shall restrict ourselves
to the description of a single experiment chosen at random from the
laboratory journal:

The muscles from the hind legs of a rat were carefully minced.
0.3 g of this minced tissue were placed in a small flask A together
with 0.05 ml of ethanol amine, 40 mg of methionine and 4 ml of
Mcllvaine’s phosphate-citrate buffer (pH ~ 7.1). Another flask
B contained a blank consisting of exactly the same amounts of
tissue etc., but no methionine. They were both incubated for 14
'hours at 37° C. The reaction mixture was then deproteinated with
16 ml 10 % trichloroacetic acid and the precipitate removed by
centrifugation. Choline determinations were carried out by the
methods described above. To avoid unnecessary tabulation we
give only the increase in choline content in flask A compared to
that in flask B expressed in per cent. The results were:

hlARENZi and Cardini method 20 %

Beattie method 19.1%

Eoman method 22 %

(Double analyses were always carried out.)

A^Tien methionine was replaced by methionine sulphoxide in
the same amoimt in the above experiment, all other data remaining
unchanged (and also the muscle tissues coming from the same
animal), the following results were obtained;

Maeenzi and CARniNi method 10 %

Beattie method 8 %

Roman method 9.5 %

A series of experiments was carried out in which the relative
amounts of muscle tissue, ethanol amine and methionine or
methionine sulphoxide were systematically varied. The results
were always qualitatively the same as in the example quoted above.
It was also found that the pH dependence of the reaction was
the same both for methionine and methionine sulphoxide.

Work on liver tissue gave similar results.

BIOBENESIS OF CHOLINE AND CREATINE.

299

Synthesis of creatinine. A typical experiment was as follows:
Flask A contained;

0.2 g minced muscle tissue
12 mg guanidine acetic acid
50 mg metliionine
4 ml phosphate buffer (pH 6.8).

Flask B contained exactly the same amounts of the various
components, but no methionine. Both flasks were incubated for
14 hours at 38° C. After tliis period 4 ml 20 % trichloroacetic
acid were added and after centrifugation 2 ml of the liquid were
removed for analysis according to the method of Lied and
Zacherl.

In order to apply the method of Benedict-Beiire-Langley
Evans-Lehnartz an experiment identical with the one just
described was carried out, iritli the following modifications. After
the incubation period 4 ml 20 % trichloroacetic acid were added
together with 1 ml 10 % hydrochloric acid. The mixture was
centrifuged after 1 hour, and 5 ml of the supernatant liquid were
autoclaved at 130° C for 30 minutes. After autoclaving the liquid
was cooled down and after addition of some methyl red indicator
it was exactly neutralised with NaOH. By addition of water the
volume was increased to 11 ml. 10 ml of a 6% dinitrobenzoate
solution were now added together with 10 ml of a 20 % sodium
acetate solution and 1 ml of a 2.5 n NaOH solution. After shaking
the flasks were left standing for 5 minutes. Their content was
then quantitatively transferred to 50 ml measuring flasks which
were filled up to the mark vnth water. The colour was measured
in a Pulfrich photometer.

Similar experiments were carried out vdth methionine replaced
by the same amount of methionine sulphoxide.

Expressing the results, for the sake of brevity, as the percentage
increase in total creatinine in flask A compared to that in flask B,
we found the following results:

Lieb and Zacherl method . ,
BENEDiCT-BEHRE-etc. method

Alcthionine

10 %
10.5 %

Methionine

sulphoxide

5 %
5.0 %

As regards the pH dependence of the process the results for
methionine sulphoxide were completely analogous to those pre-
viously obtained for methionine. Their main feature was a fairly
broad optimum somewhat above the neutral point.

300

QUNNAR STEENSHOLT.

Work on liver tissue gave analogous results, but further nu-
merical details are probably of little or no general interest and are
therefore omitted.

Commentary.

The data presented above show that the three methods used
for choline determination yield results as regards the methyla-
tion of ethanol amine which are in substantial agreement with
one another. The conclusions of a previous paper (Steensholt
1945 a) are therefore further corroborated. The experiments also
show that methionine sulphoxide is actually capable of acting
as a methyl donator, but that its efficiency in this respect is con-
siderably less than that of methionine.

The writer is glad to express his best thanks to Prof. E. Ege
for his support and generous hospitabty.

Summary.

It is found that methionine sulphoxide is capable of acring as
methyl donator in the methylation of ethanol amine to choline
and of guanidine acetic acid to creatine. The conclusions of
this and of a previous paper are corroborated by application of
different methods for the determination of choline and creatine.

Beferences.

Beattie, E. J. E., Biochem. J. 1936. 30. 1554.

Benedict, St. E., and J. A. Behke, J. biol. Chem. 1936. 114. 515.
Cazeneuve, a.. Analyst. 1900. 25. 331.

Guggenheim, M., Die biogenen Amine, Basel und New York 1940.
Langlev, W. j., and M. Evans, J. biol. Chem. 1936. 115. 333.
Knorr, L. Ber. dtsch. chem. Ges. 1897 . 30. 909.

Lehnartz, E., Hoppe-Seyl. Z. 1941. 271. 265.

Lieb, H., and M. K. Zacherl, Ibidem 1934. 223. 169.

— , — , Wien. klin. Wschr. 1934. 17. 471.

Marenzi, a. D., and 6. E. Cardini, J. biol. Chem. 1943. 147. 363.
Nencki, M., and N. Sieber, J. prakt. Chem. 1878. 17. 477.

Eoman, W., Biochem. Z. 1930. 219. 218.

Steensholt, G. Acta physiol, scand. 1945. 10. 320. 333.

DU ViGNEAUD, V., J. P. CHANDLER, M. CoHN, and G. B. BrOWN,
J. biol. Chem. 1940. 134. 787.

Werle, E., Die Chemie 1943. 56. 141.

From the Physiological Department, Karolinska Institutet, Stockholm.

The Splaiiclmic Efferent Outflow of Impulses
in the Light of Ergotamine Action.

BO GERNANDT and YNGVE ZOTTERMAN.

Received 22 February 1946,

In the course of previous research on the efferent impulse
traffic in the splanchnic nerve of the cat (Gernandt, Liljestrand
and Zotterman 1946) we tested in a few preparations the effect
of ergotamine in doses of 0.05 to 0.1 mg per kg body weight.
As ergotamine injected in these amounts produces a definite
change in the response to asphyxiation, we considered that it
might be worth while to make a study of the action of this drug,
as we now had the possibility of adding some new facts concerning
the discharge of efferent impulses in the splanchnic nerve which
might serve to elucidate the rather puzzling mode of action of
ergotamine.

HetlKuls.

All the experiments have been made upon cats anaesthetized with
O.oc g chloralose per kg body -weight.

The splanchnic nerve was exposed from the back and the efferent
discharge in the nerve was recorded according to the, method pre-
viously described.

The electric stimulation of the peripheral end of the nerve was
performed by different means. Mostly we used break-induction shocks
given with Baltzar’s apparatus in connection with a coreless induction
apparatus as described by Ludwig (1889). With this apparatus the
frequency of stimulation was varied from 30 per minute to 32 per
second without changing the strength or duration of the separate
stimulus. This strength was varied inversely with the total resistance

302

BO GERNANDT AND YKGVE ZOTTERMAN.

of the primary circuit. In a few of our earlier experiments we used
a thyratron oscillator driven from the AC mains. This stimulator
provided condenser discharges of varying frequency and rate of dis-
charge. Finally, we used a battery-driven stimulator by means of
which rectangularly shaped shocks of various frequency could be
applied to the nerve. The duration of the shock could be varied from
0.2 to 10 msec, and the strength of the current applied to the nerve
was directly controlled by a milliamperemeter.^

The arterial blood pressure was recorded with a Hg-manometer
from the femoral or the left carotid artery. Injections were made
through the femoral vein. The ergotamine substance used throughout
our experiments was the tartrate of ergotamine, Gynergen, manufac-
tured by Sandoz A.6. and the adrenaline used has been the Exadrin
produced by AB Astra.

Results.

The Effect of Ergotamine on the Efferent Discharge
of Impulses in. the Splanchnic Nerve.

Asphyxiation. As has been previously described (Gernakdt,
Liljestratstd and Zotterman 1946), asphyxia causes a gradual
increase in the efferent discharge of action potentials in the
splanchnic nerve. This discharge increases in jerky steps and gen-
erally reaches a maximum after about 46 to 60 seconds. After
an intravenous injection of ergotamine in doses of 0.05 to 0.1
mg per kg body weight the electric response of the nerve often
started earlier, and generally reached a higher maximum within
45 seconds. The general appearance of the electric activity is
otherwise very much the same in both cases (fig. 1). It is, however,
of great interest to relate this activity to the changes in the
arterial blood pressure appearing under these conditions. Whereas
normally, asphyxia is followed by a pronounced rise in blood
pressure (see fig. 2), there is a fall of blood pressure after admin-
istration of ergotamine. Concomitantly with the more rapid
increase of efferent impulses in the splanchnic nerve, the blood
pressure fell after these relatively moderate doses of ergotamine.
This reversal of the effect upon the blood pressure after a dose
of 0.05 mg per kg body w'eight generally vanished after about
45 minutes, but could be reestablished after a repeated dose. For
the further analysis of the effect of asphyxia we tested the effect
of an excess of carbon dioxide as well as of mixtures low in oxygen

* We are indebted to T. Helme, M. A., who has constructed tins device, for
his kindness in placing it at our disposal.

303

SPLANCHNIC EFFERENT OUTFLOW.

aaaaaaaaaaaaaaaaaaaaaa/

/\a/\aaAa7\7^^

B i -r*i" niinrtrrtr*- Vyi i 'r* ■— “** y f<i * « a~tu r-r j.-

[pjvvw

vwvv^

a7ww^^

E I..I u I I mi . Ill . I III 1 1 ' ~~ 1 1 1 1 I • I 'I ~i^i II I

Fig. 1. Cat 2.8 kg. Action potentials from central end of the splanohnie nerve.
A. Control, BP 140 mm Hg. B. Art. respiration stopped for 45 seconds, BP 200 mm
Hg. G. Control after 0.05 mg ergotamino per kg, BP 150 mm Hg. D. Ditto; art.
respiration stopped for 45 seconds. BP now sunk to 110 mm Hg. E. Art. respiration
again for 20 seconds, BP rpse to 130 mm Hg. The small, very quickly conducted
spikes derive from afferent fibres firing from the peripheral out end of the nerve.

Time 50 cycles per second.

on tlie cat, the blood pressure being recorded from the femoral
artery, and tbe splanchnic innervation left intact. Ten per cent
carbon dioxide in air produced a very marked fall in the blood
pressure (fig. 3).

Gas mixtures containing less than 8 % oxygen in nitrogen also
produced a fall in pressure, but the action was not so pronounced
as that of carbon dioxide. Thus it seems as if the depressing effect

304

BO GERNA^'DT AND YNGVE ZOTTERJIAN.

A. b.

r*ioo

Fig. 2. Cat 2.S kg. Chloralose anaesthesia. This figure is to be read with the electro-
neurograms of fig. 1. A. The first arrow marks the exposure of the record in fig.
1 A. The second arrow marks the stopping of the artificial respiration, the third
arrow corresponds to fig. 1 B. and the fourth arrow marks the recommencement
of art. respiration. B. The same animal after 0.05 mg ergotamine per kg body-
weight intraveno\isly. The first arrow refers to fig 1 C. The second arrow marks
the stopping of art. respiration; the third arrow refers to fig 1 D; the fourth arrow
marks the recommencement of art. respiration; the fifth arrow refers to fig. 1 E.

Fig. 3. Cat 2.G kg. After intravenous injection of 0.05 mg ergotamine per kg body
weight. A. Between the arrows artificial ventilation with 10.8 % CO. in air. B.
Between arrows art. respiration with 7.S % Oj, in N..

of asphyxiation on the blood pressure after ergotamine in these
doses is due principally to the accumulation of carbon dioxide.
This effect was not influenced by denervating the carotid sinus
or by severing the vagi in the neck.

The Effect of Electric Stimulation of the Splanchnic Nerve.

When the intact nerve is stimulated the strength of stimulus
required to elicit a reflex response from the abdominal muscles
is far below the strength necessary to induce a rise in blood

SPLANCHNIC EFFERENT OUTFLOW'.

305

Fig. 4. Cat 2.7 kg. Electric stimulation with break shook-s of constant strength'
but varying frequency to the peripheral end of tho splanolinic nerve. Between
A and B tho Xllth and Xlllth spinal roots were severed. Tho numbers below
the signal give tho frequeney of shocks per second.

pressure from the peripljoral end of the splanchnic nerve. When
using rectangular currents it was found that the abdominal re-
flexes were elicited by currents of low strength and short dura-
tion, while a rise in blood pressure could not bo elicited from the
peripheral end unless the stimuli were of long duration and at
least three times that strength.

This result was of course to be expected. The afferent fibres
concerned in the abdominal muscle reflex undoubtedly consist
of class A fibres, while the constrictor fibres as well as the secre-
tory fibres running to tho adrenal glands belong to class B,
according to the nomenclature of Erlanger and Gasser.

When the splanchnic nerve was severed above the Xllth
thoracic root and stimulated, wo found that low frequency
stimulation caused a slight drop in the arterial blood pressure.
At higher frequencies, above 8 per second, the effect is less,
and as the frequency of stimulation increases, the fall is followed
by a slight rise in pressure. In these cases, however, there are
one or two afferent connections vith the spinal cord, via the Xllth
or Xlllth spinal roots. When these connections were cut, the
same stimulation which previously produced a fall in blood
pressure now elicited a rise (Fig. 4). The fall in blood pressure
was thus obviously due to the stimulation of afferent fibres
running into the Xllth and possibly also the Xlllth thoracic
roots, which fibres caused a depressory reflex action in conformity
with Bradford (1889).

After ergotamine the rise of blood pressure was definitely
delayed. After small doses the latency was prolonged for a few
seconds, but after a large dose of 0. 5 — 1 mg per kg body weight
this retardation was very much pronounced, the latency being
prolonged from normally 1 sec. to 25 seconds when 32 induction

20 — h60215. Acta phys, Scandinav. V ol, 11.

806

BO 6ERNANDT AND YNGVE ZOTTEBMAN.

fi, Fb Fib Fsi ^s^ Fib F8 Fi

Fig. 6. Cat 3.2 kg. Stimulation as in previous figure. A. before, B. after 1 mg
ergotamine per kg intravenously. Note the increase in latency and the retardation
of the response after ergotamine.

shocks per second were applied. The pressure rose much more
slowly and the duration of the rise was very much prolonged
(fig. 5 A and B).

The Besponse to Adrenaline.

After moderate doses of ergotamine, 0.05 — 0.1 mg per kg body
weight, small amounts of adrenaline in doses of 0.25 — 1 fj, g
injected intravenously produced a more accentuated rise in
the arterial pressure in accordance with the observation made
by Euler and Schmiterlow (1944). The return to the original
pressure level also appeared to be somewhat slower. The rise
in pressure was accompanied by a diminution of the efferent
discharge of the splanchnic nerve (see fig. 1 C). After larger
doses of ergotamine, 0.5 — 1 mg per kg, the effect of small doses
of adrenaline was considerably delayed, . and the rate of rise
much reduced (fig. 6).

, 30"

^ 30“ aJr. p .

Fig. 6. Cat 3.6 kg. A. Control. The effect of 2 ng adrenaline. B. After 1 mg ergo-
tamine per kg intravenously. Note the delay in response to the intravenous in-
j ection of adrenaline as well as to the stimulation of the peripheral splanchnic nerve.

A reversal of the adrenaline effect we have seen in cats with
intact brain and under chloralose anaesthesia only after larger

SPI-AXCHNIC EFFERENT OUTFLOW.

307

doses tlian 2 mg per kg body weight. Even in these conditions
the phenomenon is by no means of constant occurrence. Generally,
the reserved effect is observed when the level of the blood pres-
sure is high. AVhen the level is below 90 mm Hg we have not
seen any reversal, in conformity with Cannon and Lyman
(1913). After pithing the cat the reversal seemed to occur more
constantly and seemed to be less dependent upon the level of
the blood pressure. In the pithed cat, the reversal was generally
seen after 1 mg ergotamine per kg (Fig. 7). With this dose of
ergotamine we never obtained any reversal in a cat w'ith intact
brain. After larger doses of ergotamine
there was generally a ver}'^ marked
increase in the vagal tonus and we
therefore often severed the vagi. This
procedure did not lead to any change
in the response to adrenaline.

There is no reason to expect that
adrenaline should exert any direct
action upon the vasomotor centres.

We also found that the discharge
of impulses in the splanchnic nerve
changed only in response to the
change of blood pressure. Thus w’hen
adrenaline caused a rise in blood
pressure the discharge was diminished,
while the reversed effect of adrenaline
after ergotamine led to an increase in
this discharge. These changes were
very closely related to the change in pressrire and not to the
amount of adrenaline given. When for example after a large
dose of ergotamine (1 mg per kg) there was a very slight rise in
blood pressure in response to 0.1 mg adrenaline, the splanchnic
discharge was very faintly reduced. Before ergotamine, ■when
the same dose of adrenaline elicited a very high blood pressure,
the discharge was entirely wiped out.

After decapitation, when the spinal cord was transsected at
C II, there was sometimes a very pronounced increase of the
splanchnic discharge (see fig. 8 C). This increased discharge,
which had the usual appearance, was always accompanied by
a very marked rise in blood pressure to 200 mm Hg. During
10 — 20 minutes this discharge gradually diminished, and there

. , .... , ji

Fig. 7. Pithed cnt. Imgcrgotn-
lainc per kg body weight intra-
venou.sly. The first signal mark-
ing refers to faradic stimulation
of tlio periplicral end of the
splanchnic nerve with 32 shocks
per second for 30 seconds. There-
after intravenous injection of
0.1 mg adrenaline. Note the
prompt start of the reversal.

308 BO GERXAXDT AXD VXGVE ZOTTEBJtAX.

vwxaaaaaaaaaaaaaaAA/

VWWXAAAAAAAAAAAAA/V

B - m -it ^i>i^i< i i ^,<iii, n ..n i ..iii| Mni it i lu .

AAAAAAAAAAAAAAAAA/Vi^

VAAAAAAAAAAAAAAAAAAAA

VNAAAAAAAAAAAAA/^^

Fig, 8. Cat 2A kg. Action potentials from central end of the splanchnic nerve.
Distance between off-leads 2 mm. A. Control. BP 130 mm Hg. S. After 6 //g
adrenaline, BP 190 mm Hg. The small diphasic and very rapidly conducted spike
potentials seen are injury potentials from afferent fibres. C. Immediately after
decapitation, BP 180 mm Hg. D. 20 minutes later BP 60 mm Hg. S. Artificial
re.spiration stopped for 45 seconds. BP rose from 60 to 80 mm Hg. Time 50 cycles

per second.

was a parallel fall in blood pressure until tbe pressure as well
as tbe discharge was stabilized at a lower level (see fig. 8 D).
This phenomenon we are inclined to ascribe to a traumatic effect
which is gradually diminished.

In the spinal animal asphyxiation elicited a very marked in-
crease in the discharge, and this increase was quite independent
of whether the blood pressure rose or fell, the latter being the

efferent outflow.

309

SPLANCHNIC

VVV\AAAAA/WW\AAA^^

A . 1 '

\AAA/VW\^^

XAAAAAAAAAAAAAAAAAAAA/'

aaaaaaaaaaaaaaaaaaaaa/

[

Fig. 9. Action, potentials from the central end of the splanchnic nerve of the cat.
Same animal as in fig. 8. A, After 1 mg ergotamine per kg, BP 90 mm Hg. B.
After decapitation, BP 110. 0. Art. respiration stopped for 45 seconds. BP fall
to 80 mm Hg. B. Art. respiration again for 30 seconds, BP rose to 100 mm Hg.

case after ergotamine in doses of 0.05 — ^2 mg per kg body •weigtb.
(Kg. 9). _

In tbe spinal cat we could thus not observe that ergotamine
produced any other change in the behaviour of the pre-
ganglionic response to asphyxia than that the discharge V'as
somewhat increased, which we presume to be due to the fall
in pressure.

The splanchnic outflow of efferent action potentials was
once observed 10 minutes after the heart had stopped (see
fig. 10 A). This activity was found to originate within the

spmal cord, as it was completely abolished by cutting the spinal
roots (fig. 10 B). or

310

BO GERSANDT AND YNGVE ZOTTERMAN.

WWVWWW\AAAAAAAA/V'

VVVWWWWXAAAAAAAAAAy

Fig. 10. Action potentials from the central end of the cat’s splanchnic nerve.
Upper record shows the efferent outflow of impulses 10 minutes after the death
of the animal. Lower record immediately after severing the nerve centrally to
the off-leads. Distance between electrodes 2 mm. The rapid small spikes seen
in the lower curve are antidromic spike potentials from afferent fibres firing from
the freshly cut part of the nerve centrally to the electrodes.

Discussion.

The increased rapidity in the response of the vasomotor centres
to asphyxia as observed after even moderate doses of ergotamine
is a remarkable fact when correlated with the fact that the action
of the increased B-fibre activity in the splanchnic nerve now
coincides with a fall in blood pressure. This fall in blood pressure
has been clearly proved to be due principally to the accumula-
tion of carbon dioxide, but this fact alone cannot explain the
reversed action upon the blood pressure.

According to Eothlin (1923) the depressor reflexes are di-
minished after moderate doses of ergotamine, and v. Euler
and ScHMiTEELdw (1944) have produced evidence for the view
that this action of ergotamine consists in a centrally located
inhibition of the action of the baroceptive impulses upon the
vasomotor centre. Thus the readier response of the splanchnic
efferent fibres after ergotamine would depend upon the elimina-
tion of the continuous inflow of these inhibitory impulses from
the aortic and carotid sinus regions. But in order to understand
the fact that the augmented fibre activity in the nerve after
ergotamine produces a fall in blood pressure, we have to discuss
the following possible explanations.

1) The impulses observed derive both from vasoconstrictor
and vasodilator fibres, but the action of the latter predominates,
as the former are inhibited peripherally by ergotamine.

SPLANCIIJCIC EFFBREM OUTFLOW.

311

2) The vasoconstrictor fibre impulses as well as the increted
adrenaline elicit a reversed action directly upon the blood vessels
after ergotamine.

The supposition of the existence of specific vasodilator fibres
in the splanchnic nerve is based upon the old observations made
by Bradfobd (1889) and Johanssox (1890), who found that
very slow frequency stimulation (30 — 60 stimuli per minute)
of the peripheral end of the splanchnic nerve of the dog often
caused a small fall in the arterial blood pressure, while more
frequent stimuli raised the pressure. In the cat, hoAvever, we
have not been able to produce a fall in pressure by low frequency
stimulation of the peripheral end of the splanchnic nerve provided
all the afferent connections of this nerve with the spinal cord
were severed. A fall in blood pressure produced by electric stimu-
lation of the peripheral end of the splanchnic nerve we have
observed only after large doses of ergotamine (1 mg per kg
body weight), and this only after the cat had been pithed.
It Avas only under these conditions that we found a reversed
effect upon the blood pressure of intravenous adrenaline injec-
tions. After 2 mg per kg body Aveight the adrenaline reversal
was generally observed in cats AAith the brain intact. In the pithed
cat the fall in pressure starts as promptly in response to faradic
stimulation as does the rise under normal conditions. In the latter
case the immediate rise is apparently caused by the direct vaso-
constrictor response, which is folloAA'ed by a secondary rise caused
by the adrenaline liberated from the adrenal gland. The prompt
fall in the blood pressure produced by stimulation of the splanch-
nic nerve in the pithed cat under large doses of ergotamine occurs
with as short a latency as does normally the Amsoconstrictor
response. Thus it seems as if the first part of the fall Avere due
to a direct action upon the abdominal vessels, in conformity
Avith Dale (1913). Dale interpreted this reversed action in the
pithed cat as due to the activity of specific vasodilators Avhich
are stimulated concomittantly A\'ith the vasoconstrictors, the
latter being blocked peripherally by the ergotamine. This fall
cannot be produced by using AA^eaker stimuli than those Avhich
normally produce the rise, and our experiments AA’ith stimuli
of various duration and frequency have not yielded any evidence
for the view that constrictor and dilator fibres are different in
respect of their reaction to electrical stimulation.

If the constrictor fibres are blocked peripherally by, ergota-

312

BO GERNANDT AND YNGVE ZOTTERMAN.

mine, the action of the mixed volleys of constrictor and dilator
impulses must lead to a fall in blood pressure. With small doses
of ergotamine there is a retardation of the rise in blood pressure
in response to faradic stimulation of the splanchnic nerve, and this
retardation becomes more pronounced as the amount of ergotamine
is increased. The important point to remember, however, is that
the effect of stimulation as well as the effect of intravenous in-
jection of adrenaline was not reversed until the cat was pithed,
or the amount of ergotamine administered was raised to about
2 mg per kg body weight.

Let us now consider the second alternative. We have already
mentioned that when there is a reversed reaction to stimulation
of the splanchnic nerve after ergotamine it starts very promptl3\
The same holds true for the reaction to intravenous injections
of adrenaline. If we accepted the view of specific dilator fibres
we were thus compelled to assume that these fibres were adrenergic
in nature. Bulbring and Burn (1935) have shown that the leg
muscles of the cat are supplied only with adrenergic vasodilators.
We should thus accept Dale’s original view that ergotamine
inhibits the sympathetic motor action but rather exaggerates
the sympathetic inhibitory actions, and thus reveals the vaso-
dilator fibre activity. The abolishment of the motor action is
proved to be peripheral, but the inhibitory action — in this case the
vasodilator action — might be elicited purely peripherally, i. e. by
the transmitter substance liberated and by the adrenaline secreted
from the adrenals. Accepting the theory of sympathin E and I
advanced by Cannon and Rosenblueth, the increasing massive
outflow of impulses in the sympathetic efferent fibres following
upon asphj^ia must be assumed to produce a peripheral libera-
tion of sympathin E and I as well as an incretion of adrenaline
from the adrenal glands. Normally, this leads to a vasoconstriction
and rise in blood pressure. After ergotamine, however, we may
assume that the stimulating action of sympathin E as well as
the motor action of adrenaline is reduced or abolished, while
the inhibitory action of sympathin I and adrenaline is exaggerated.
The increased outflow of efferent impulses will thus produce
a reversed effect. The action potentials recorded from the splanch-
nic nerve must thus be looked upon as being built up by axon
potentials deriving from efferent fibres of different functions; a)
vaso-constrictor fibres, b) adreno-secretory fibres, c) vaso-dilator
fibres (?), d) inhibitory fibres to the intestines. In the peripheral

SPLAXCIIXIC EFFEUEXT OUTFLOW.

313

endings of the first group sympatliin E is jwobably liberated.
The adreno-secretory fibres are obviously cholinergic. The pre-
sumed vasodilator fibres and the inhibitory visceral fibres produce
sympathin I. Now the synipathin I liberated is norinally over-
powered in its action upon the blood pressure by the combined
action of sympathin E and adrenaline, but after crgotamine the
sympatliin I activity will be revealed.

These assumptions seem to agree very well with the experience
from our experiments in which we directly stimulated the periph-
eral end of the splanchnic nerve. IVhen applied with adequate
strength and frequency the faradic stimulation normally produced
a rapid rise in pressure. After ergotamine the rise in blood pressure
is delayed and the rise is retarded. This delay in the onset of the
rise in pressure is gradually prolonged with increasing doses of
ergotamine, until after 1 — 2 mg per kg the stimulation at once
causes a fall in pressure. The responses to adrenaline under these
conditions show very much the same course. First there is an
exaggerated response due to the iniiibition of the depre.ssor
reflexes. As the ergotamine doses are increased there is an in-
creasing delay in the response until a point is reached where
adrenaline causes the well-known reversal. So far everything
seems to fit the scheme.

We have, however, to consider the fact that asphyxia caused
a reversed effect upon the blood pressure even after relatively
moderate doses of ergotamin. This reversed effect of asphyxia
never failed after a dose of 0.06 rag per kg. The fall in pressure
is vdth tliis dose often not immediate, but is preceded by a small
rise. With increased doses of ergotamine the preliminary rise is
reduced, however, and finally there is only a direct fall in pressure.

How are Ave to reconcile this behaviour in response to asphyxia
with our above-mentioned vicAV on the ergotamine action? After
0.05 mg or ergotamine per kg the rise in pressure in response
to a small amount of adrenaline is rather exaggerated, but the
effect of asphyxia is reversed. This does not seem to fit in AA'ith
the Anew referred to. We must, hoAvever, bear in mind that it
IS not necessary, or rather it is not correct, to assume that the
effect of intraAmnous injection of adrenaline AA'ould haAm the same
action upon the blood pressure as has the increased outfloAA’ of
efferent impulses in the splanchnic ncrAm observed during asphyxi-
ation. The motor effect of this discharge can be much more liable
to the peripherally depressing action of ergotamin than is the

314 BO GERNAKDT AND TNGVE ZOTTEBMAN.

injected adrenaline. AVe liave further to consider that during
asphyxia the accumulating metabolites exert a peripheral vaso-
dilatory action of increasing strength, ■which normally is over-
■ff’helmed by the constrictor and adrenaline outflo'w. AVhen
ergotamine has •weakened the motor effect this dilating effect
is added to that of sympathin I. This addition brings the reaction
upon the blood pressure to a reversal. It is also clear that the
direct stimulation of the splanchnic nerve will produce much
more massive volleys of impulses than any reflex activity could
ever produce. Thus the constrictor effect "will "win until at last
the motor action, or let us say the sympathin E action, is de-
pressed by large doses of ergotamine. The same scheme can be
applied to adrenaline. The total abolishment of its motor action
demands large amounts of ergotamine.

The question as to whether the splanchnic nerve of the cat
contains special sympathetic dilatory fibres to the abdominal
blood vessels could not be settled in these experiments, as the
electric response recorded did not permit any detailed analysis
of its fibre contribution.

The dilator fibres, if they really exist, seem, however, to dis-
play the same properties, viz. electric stimulation, as do the
constrictor fibres. Thus the possibility of distinguishing the action
potentials of the constrictor fibres and the presumed dilator
by their shape seems rather remote.

The strongest direct evidence for the existence of specific
dilator fibres in the cat’s splanchnic nerve hitherto advanced is
Dale’s demonstration in 1913 that faradic stimulation of the
peripheral end after adequate doses of ergotamine caused a slight
fall in blood pressure in the pithed cat after removal of the
adrenal glands. AA^’e have, however, to remember that the faradic
stimulation which normally excites the constrictor and the
adreno-secretory fibres no doubt affects most of all the other
fibres, afferent fibres and inhibitory fibres to the intestinal
smooth muscles. The antidromic afferent impulses might thus
produce a dilatory action to which the effect of the sympathin
I producing inhibitory fibres might add. This would agree very
well with the assumption that adrenaline elicits motor as well
as inhibitory actions, "while the substance produced by con-
strictor impulses is pmely motor-active. It is only the motor
activity which is abolished by ergotamine. Thus adrenaline will
cause a reversal which sympathin E would not produce. The

SPLANCHNIC EFFERENT OUTFLOW.

315

effect of sympatLin E can thus be expected to be gradually
depressed with the amount of ergotamine given and sympathin
E will never cause a reversal.

We thus conclude that it seems rather likely that the increased
discharge of action potentials following upon asphyxiation before
as well as after ergotamine derives from the same fibres. They
w’ould thus derive from vasoconstrictor fibres, adreno-secretory
fibres and inhibitory fibres to the intestines. The reserved reac-
tion after ergotamine to this increased discharge thus seems to
be entirely peripherally localized. The simplest explanation
w'ould therefore be to assume that the sympathetic nerve activity
is exerted via a transmitter which has all the properties of adre-
naline. To this the adrenaline liberated from the adrenals w’ill
add ist action.

After ergotamine in sufficient doses the efferent outflow of
sjTiipathetic impulses elicited by asphy.xiation or by direct faradic
stimulation of the splanchnic nerve w'ill thus produce the same
reversed effect upon the blood pressure as does adrenaline in-
jected intravenously.

It must, however, be admitted that all the observations here
related can be understood equally well by assuming that the
sympathetic motor action is caused by Cannon’s and Rosen-
blueth’s sympathin E and the inhibitory action by sympathin
I, and that adrenaline can give rise to both actions. The efferent
outflow which normally raises the blood pressure via sympathin
E and the adrenaline increted from the adrenals will not be
able to exert this effect after ergotamine w'hich depresses' or
blocks the . action of sympathin E as well as the motor action of
the adrenaline liberated from the adrenal glands. This adrenaline
will now' exert its sympathin I activity. As Dale (1913) has
demonstrated, however, faradic stimulation of the peripheral
end of the splanchnic nerve causes a slight fall in blood pressure
after removal of the adrenal glands. As discussed above, this
effect may be due to a stimulation of afferent fibres and inhibitorj'
fibres running to the intestines. The former fibre would cause
a peripheral liberation of acetylcholine and the latter of sympa-
thin I, w'hich might affect the blood vessels as w'ell as the intestinal
muscular coat.

When we assume that the sympathetic nerve activity peripher-
allj' is mediated by a transmitter sympathin E or sympathin I,
it seems that we must reckon with the possibility of specific

316

BO GERKASBT ASD YXGVE ZOXTEKMAiSr.

sympathetic vasodilator fibres in the splanchnic nerve or assume
that the inhibitory fibres to the intestines exert a dilatory effect
in order to understand the reversed effect after ergotamine.
If we assume that the constrictor fibre activity is mediated by
adrenaline, the assumjjtion of specific dilatatory fibres in the
splanchnic nerve is unnecessary.

Thus the problem as we see it is not whether there exist dilator
fibres in the splanchnic nerve which may be excited electrically,
but whether there are specific dilator fibres with no other func-
tion.

In such a case one would expect that a rapid rise of the blood
pressure as produced by an adequate dose of adrenaline in the
normal cat would via the depressor reflexes lead to an outflow
of dilator impulses in the splanchnic nerve. In conformity with
Adrian, Bronk and Phillips 1932 we found that the effect
was a complete inliibition of the ordinary outflow and nothing else.

Thus, our study of the splanchnic impulses has not afforded
any direct evidence for the existence of specific vasodilator fibres
in the splanchnic nerve of the cat, and as the outcome of the above
discussion we must conclude that their existence seems to us
to be still more uncertain than before.

Summary.

For an analysis of the action of ergotamine upon the arterial
blood pressure of the cat we have recorded the action potentials
from efferent fibres of the splanchnic nerve.

When ergotamine is given in a moderate dose of 0.05 mg per
kg body weight, asphyxiation or the inhalation of air rich in
carbon dioxide produces a fall instead of the usual rise in blood
pressure. The electric response, however, differs from that to
ergotamine only in that the response now starts earlier and is
more accentuated.

After a pithing of the brain and the medulla oblongata the
efferent outflow in the splanchnic nerve reacts to asphyxiation
and carbon dioxide in a way very similar to the reaction in the
intact cat, but the response is more rapid and more pronounced
than in the intact animal. Ergotamine does not produce any change
whatever in the splanchnic efferent outflow in the spinal cat.

This shows that inhibitory influences upon the spinal vaso-
motor centres are exerted from higher centres, which influence

8PLANCI1NIC EFFERENT OUTFLOW.

317

is abolished by ergotamine even in moderate doses, as has been
shown by Eothlin, Weight and Euler and Schmiterloav.

When giving increasing doses of ergotamine it was found that
the effect of faradic stimulation of tlie peripheral end of the
splanchnic nerve, wliich normally causes a rapid rise in the blood
pressure, became more and more delayed and depressed, until
after large doses the effect was reversed. At his stage the adrena-
line effect was also reversed.

These observations have been discussed in the light of the
prevailing theories on the nature of peripheral sympathetic effect.

It was found that the simplest explanation would be to assume
that the sympathetic nerve activity leads to a peripheral libera-
tion of adrenaline. The phenomena observed under ergotamine
may, however, be equally well understood by assuming that
the constrictor excitation is mediated by sympathin E. In such
case we may assume that the vasodilation produced by direct
stimulation of the splanchnic nerve is due either to the action
of antidromic afferent impulses or to the liberation of sympatliin
I in fibres running to the intestinal walls, or to a combination
of both these influences. Thus even in this scheme the supposition
of specific sympathetic vaso-dilator fibres in the splanchnic
nerve of the cat does not seem to be necessary.

The recording of the action potentials from the splanchnic
nerve has not yielded any proof of any activity of vasodilatory
fibres either before or after ergotamine.

The expenses for this research have been defrayed by a grant
from AB Astra.

References.

Adrian, E. D., D. Bbonk, and G. Phillips, J. Physiol. 1932. 74 . 115.
Bradford, J. R., J. Physiol. 1889. 10 . 358.

Bulbring, E., and .1. H. Burn, Ibidem. 1935. 83 . 483.

Cannon, W. B., and E. Lyman, Amer. J. Physiol. 1913. 31 . 367.
Gannon, W. B., and A. Rosenblueth, Ibidem, 1933. 104 . 557.

Dale, H. H., Ibidem. 1906. 34 . 163.

Dale, H. H., Ibidem. 1913. 46 . 291.

Euler, U. S. v., and 0. Schmiterlow, Acta Physiol. Scand. 1944,

8. 122 .

Gernandt, B., G. Liljestrand, and Y. Zotterman, Ibidem. 1946.
Johansson, J. E., Sv. Vet. Akad:s Handl. 1890, B 16. Afd. IV. Suppl.
n;o 4.

Ludwig, C., Arch. Anat. Physiol., Lpz. 1889. p. 275.

Rothlin, E., Arch. int. Pharmacodyn. 1923. 27 . 459.

Wright, S., J. Physiol. 1930. 69 . 331,

From Universitetets Biokemiske Institut, Copenhagen.

On tlie Effect of Some Pigments and
Redox Systems on the Respiration
of Animal Tissue.

GUNNAR STEENSHOLT.

Received 23 February 1946.

In recent years considerable attention has been given to the
study of the effect of certain naturally occurring pigments and
redox systems on the respiration of animal and other tissue, and
a considerable literature has grown up around this problem.
It suffices here to recall the work done on pyocyanine, toxo-
flavine, echinochrome, haUachrome, and some naphtoquinones
like juglon and lawson. We cannot here discuss in detail all the
interesting facts which have been brought to light in this field.
A good review of the whole subject has been given recently by
Stern (1939), and the reader is referred to this standard work for
all references. However, much still remains to be done in this
field; it is seen from a survey of the literature that only a few
scattered substances have been investigated so far. In the present
note a report is given on certain results obtained some time ago
by the present writer in the study of substances, the investigation
of which does not seem to have been reported in the literature,
though for some of them the desirability of such an investigation
has already been pointed out.

The first of these compounds is phthiocol, the yellow pigment of
Bact. tuberculosis;

CH,

for which no respiration experiments have been carried out so far.

BESPIKATIOK OP ANIMAL TISSUE.

319

Similar work was then done on 2-raethyl-l,4-naphtoquinone,
which is not kno-^m to occur as a natural pigment in living cells,
but is known for its vitamin K activit 5 ^•

The second class of substances studied in tlu's paper is formed
by some compounds closely related to pyocyanine, namely the
following;

H I

Xy \/

• OCH,

/\/\/"\

a-oxy-phenazinc

a-methoxy-phenazine

N-ethyl-a-oxy-phenazine

Pyocyanine itself is N-methyl-a-oxy-pheiiazine.

320

flUKNAR SXEEKSHOLT.

Finally we have studied some anthocyans in their effect on
tissue respiration, namely cyanine, which is a diglucoside of- the
anthocyanidine cyanidine; peonine, the diglucoside of peonidine;
and finally malvine, which derives from syringidine and contains
two molecules of glucose. The complicated structural formulae of
these compounds are omitted here. For anthocyans and antho-
cyanidines no respiration experiments seem to have been carried
out so far. Their redox potentials have not been determined, and
practically nothing is known about their proper biological func-
tion. Reichel states that anthocyans act as hydrogen acceptors
in Thunberg experiments with purified liver aldehydrase (Rei-
chel (1937), Reichel and Kohle (1935)), but this hitherto rather
isolated observation does not lead us much further towards an
understanding of the physiological functions of these compoimds.
Nevertheless it seems of interest to carry out some respiration
experiments with these substances and some results in this direc-
tion will be reported in the present paper. Unfortunately work on
anthocyanidines could not be included in this note, but it is hoped
to return to this subject later on.

Experimental Part.

Biological material. We have worked with kidney, liver and testis
tissue from rats. The animals were from 4 to 8 months old, and had been
kept on a diet believed to' be sufficient in all respects. The animals were
killed by decapitation and the organs to be investigated were removed
immediately after death.

Substrates. Of the substrates investigated the naphtoquinones and the
phenazine derivatives were synthesised by the present writer by known
methods (see Stern, 1939, for literature). The anthocyans were ob-
tained from the laboratory of Prof. P. Karreb in Zurich, and the writer
is glad to express his best thanks for this valuable gift.

Experimental procedure. The Barcroft-Warburg manometric method
was employed. The tissue, either in a finely divided state or in the form
of slices, was placed in the main chamber of a 10 ml respiration vessel
together with 2 ml of Ringer’s solution. The central chamber contained
0.2 ml of 5 % KOH for absorption of carbon dioxide. The side chamber
contained an aqueous solution or suspension of the substance, the effect
of which on tissue respiration was to be examined. We have used in all
our experiments solutions or suspensions which were M/25, M/50 and
M/100 in the substance to be investigated. However, we quote below
only the results for the M/50 solution, the results in the two other cases
being quite similar from the qualitative point of view. The thermostat
was kept at 38°C. After equilibration for 20 minutes the oxygen con-

BESPIEATION OF ANIMAL TISSUE.

321

sumption was measured for a first period of usually 20 minutes. The
solution in the side vessel was then tipped into the main chamber,
and the observations were continued for a second period of equal length.
This method suffers admittedly from the defect that the partial pressure
of carbon dioxide is held at approximately zero, which is a rather un-
physiological condition. The great advantage of the method is its
extreme simplicity and the identity of the tissue fragments both in
the experiment itself and in the controls. It has previously been success-
fully applied by Feiedheim (1931) in his work on the effect of pyocya-
nine on the respiration of normal and tumour tissue. The tissue frag-
ments were always introduced into the previously warmed respiration
vessel within a period of less than 6 minutes from the death of the
animal. The oxygen consumption was found to be strictly proportional
to the time for at least 75 minutes. The deviations to be expected with
exclusion of bacterial growth are in the direction of decreasing oxygen
consumption. Hence these deviations cannot exaggerate any observed
increase in respiration.

As Ringer’s solution we used both ordinary bicarbonate Ringer and
phosphate Ringer. In the experiments in which sugar was added a con-
centration of 0.2 % glucose was used.

Results. To avoid superfluous and tedious tabulation of nu-
merical details we have expressed all our results ip. the following
form: The most probable values of the oxygen consumption during
the first and the second period of observation were calculated in a
familiar way by means of a diagram in which oxygen consumption
was plotted against time. The difference between these two
quantities gives the increase in respiration brought about by the
substance under investigation, and was finally expressed in per
cent of the oxygen consumption during the first period of obser-
vation.

Kidney tissue was used in all our experiments in the form of
slices cut by means of a razor blade or a sharp Icnife. Respiration
experiments were carried out in bicarbonate and phosphate
Ringer both with and without glucose, but no effect whatever
on the oxygen consumption was observed for any of the substances
investigated. The numerical details are therefore of no general
interest and are hence omitted.

Liver tissue was prepared for the experiments by placing it on a
watch glass and carefully mincing it by means of a pair of bent
scissors. The results of the respiration experiments were: The
anthocyans were found to be completely without effect. Similarly
a-methoxy-phenazine, a-oxy-phenazine and phthiocol had no
influence whatever on oxygen consumption. 2-methyl-I,4-naphtor
quinone was hkewise without effect in bicarbonate Ringer, both

21 — i60215. Acta phys. Scandinav. Vol. 11.

322

GUNKAR STEENSHOLT.

Tvitli and witliout glucose. However, in phosphate Ringer a very
distinct effect was obtained. As an example we quote the following
results of a set of experiments, with 3 respiration vessels:

Percentage increase
in oxygen consumption

Phosphate Ringer without glucose 50 35 45

Phosphate Ringer with glucose ! . . 100 40 50

Analogous results were obtained for N-ethyl-a-oxy-phenazine.
The substance had no effect on the respiration of Hver tissue in
bicarbonate Ringer with or without glucose. In phosphate Ringer,
however, an effect of the same order of magnitude as that for 2-
methyl-l,4-naphtoquinone was found. The results of a typical
experiment were as follows:

Percentage increase
in oxygen consumption

Phosphate Ringer without glucose 25 30 75

Phosphate Ringer with glucose 50 70 55

Testis tissue was prepared for the respiration experiments in the
same way as liver tissue. Here all the substances investigated were
shown to be without any effect, in spite of all efforts to the con-
trary. A tabulation of the numerical details is therefore of no
general interest and is omitted.

Finally some respiration experiments were carried out with
yeast cells. Ordinary baker’s yeast, suspended in ordinary Ringer
or phosphate Ringer, was employed throughout. Phthiocol was
found to have no effect ontherespiration of yeast cells, a-methoxy-
phenazine was likewise inactive, while for N-ethyl-a-oxy-phenazine
the effect was either zero or very small. For a-oxy-phenazine a
typical experiment was as follows:

Percentoge increase
in oxygen consumption

Ordinary Ringer 50 60 55

Phosphate Ringer ; '. 50 75 45

Similarly we obtained for 2-methyl-l,4-naphtoquinone:

Percentage increase
in oxygen consumption

Ordinary Ringer 60 80 82

Phosphate Ringer 1 50 55 72

Hence for both these compounds the effects observed were very
considerable.

RESPIRATION OF ANIMAL TISSUE.

323

Coiumcnls.

An exhaustive treatment of the results presented above is hardly
possible at the present time on account of the very complex nature
of the processes involved and our present very incomplete know-
ledge of them. Nevertheless, a brief discussion of a few points
may not bo out of place.

Phthioco! is a relatively negative redox system, and in our ex-
periments it was found to have no effect on tissue re.spiration.
The potential of 2-methyl-l,4-naphtoquinonc does not seem to be
known, but it may perhaps not be so different from that of phthio-
col. In oiu: experiments it was found to raise the oxygen uptake
of liver cells verj’’ considerably. In this connection it may be of
importance to remember that the redox potential is not the only
factor to decide whether a substance can act as carrier; its chemical
constitution, for instance, is also of importance.

The inactiidty of a-methoxy-phena;5ine can probably be well
accounted for on the basis of present knowledge of the structure
of those phenazinc compounds that form reversible redox systems
(see Stern (1939), p. 228 et scq.). N-ethyl-a-oxy-phenazine was
found to behave very much like pyocyanine itself, as one would
probably expect it to do. It is a curious phenomenon that there is
no effect except in phosphate Ringer. A somewliat similar observa-
tion was previously made by Friedheim (1934). Like pyocyanine,
the substance has no effect on cells with a perfect respiration, i. e.
exhibiting no aerobic glycolysis whatsoever (kidney). There are
some differences, however, which are perhaps not so readily
imderatood at the present time. According to Friedheim the
presence of glucose is essential for the increase in respiration ob-
served with pyocyanine, but this ivas certainly not the case in
our experiments with its ethyl homologue. Further, pyocyanine
was found to increase the oxygen uptake of testis tissue, but
N-ethyl-a-oxy-phenazine was inactive in this case.

The other results do not seem to call for much comment at the
present time. It would be interesting to continue the work, es-
pecially with anthocyanidines, and as already mentioned it is
hoped to do so in a later note.

The author is glad to express his best thanks to Prof. R. Ege
for his generous support and hospitality.

324

GDXKAR STEENSHOLT.

Summary,

Certain naplitlioqmnones, pienazine derivatives and antlio-
cyans have been subjected to respiration experiments withtbe
Babcboft-Waebtjhg technique, using Iddney, liver and testis
tissue and yeast cells. Some of the results can he fitted into our
present picture of these processes, hut certain findings appear to
he somewhat at variance with those of previous workers in this
field.

References.

Fbiedheim, E., Biochem. J. 1934. 2S. 173.

Eeiohel, L., Naturwiss. 1937. 25. 318.

— and H. Kohle, Hoppe-Seyl. Z. 1935. 236. 145, 158.
Stern, K., Biological Oxidations, The Hague 1939.

From the Laboratory for the Theory of Gymnastics,
University of Copenhagen.

Further Investigations on the
Effect of Adenosine Triphospliate and Related
Pliosphorus Compounds on Isolated
Striated Muscle Fibres.

FRITZ BUCHTHAL, ADAM DEUTSCH and G. G. ICNAPPEIS.
Received 17 Fcliriiary 1D16.

In a previous communication it has been shown that adenosine
triphosphate (ATP) in extremely small amounts (approx. 10"^ /ug)
releases contraction in normal and curarized striated muscle
fibres. Furthermore, ATP causes a reversible decrease in bire-
fringence, while inorganic triphosphate initiates contraction
without changing this property. The changes in birefringence
after application of ATP last considerably longer than the me-
chanical process of contraction.

In the present investigation the effect of ATP on striated
muscle is studied under different conditions. In view of the effect
of inosine triphosphate (ITP) in model experiments on myosin
solutions, it seemed of interest to examine the effect of this
compound on muscle fibres and at the same time to extend the
investigation to comprise other phosphorus compounds of high
bond energy. To begin with we have examined creatine phosphate
and acetyl phosphate.

Method.

The experiments were carried out on isolated fibres or small bundles
of the frog’s m. semitendinosus (Rana esculenta and Rana iempomria).
The threshold for release of contraction by the different chemical

326 FRITZ BUCHTHAIi, ADAM DEtJTSCH AND 6. G. KNAPPEIS.

stimuli was determined on small fibre bundles (2 — 8 fibres) which
were fixed by their tendon ends in two movable metal clamps on a
slide. During preparation and the experiment the fibre was kept in a
Ringer solution at 10° containing 6.7 g NaCl, 0.2 g KCI, 0.2 g anhydrous
CaOL and 0.2 g glucose per litre. The pH was kept constant at 7.3
by adding a suitable amount of NaHCOs and passing a stream of a
GOa— Oj mixture through the Ringer solution. The normal colloid
osmotic pressure was attained by supplying the Ringer solution with
3 per cent dextran (Gronwadd and Ingelman 1915). The chemical
substances were administered with a fine pipette when the amount
of liquid was O.os ml, or with a micro-pipette for liquid volumes of

Fig. 1. SIusclo chamber provided with an arrangement to turn the fibre 90° for
measuring thickness and phase difference in the same plane.

approx. 6 X 10“^ In all cases the effects of chemical stimuli were
tested in non curarized and completely curarized fibres and compared
with electrical stimulation by single current pulses and tetanic stimuli
from a thyratron stimulator. Ror the experiments on the denervated
muscles the sciatic nerves were divided under ether anaesthesia and the
animals were kept for some 3 weeks to allow degeneration to proceed
to completion.

In measurements of birefringence only isolated fibres were used,

their tendon ends being held by two pairs of silver tweezers m a
chamber at constant temperature (Rig. 1). Phase difference {F) was
determined with a Babinet compensator as in former experiments
(Btjchthal and Knappeis 1938).

When determining fibre thickness (d) we have hitherto measured
the diameter by means of an eye-piece micrometer with movable
cobweb in a plane perpendicular to that used for determination of
phase difference. In the course of other experiments we found consider-
able deviations in the circular cross section of the fibres, which may
introduce serious errors in the absolute values of birefringence. To
eliminate this the fibre was rotated through 90° by turning the two
pairs of tweezers. The fibre diameter was thus measured in the same
plane as that used for determining phase difference.

KF]?ECT OF ADENOSINE TIUPIIOSPHATE.

327

Preparation of substances.

Adenosine tripJiosphaie {ATP). The ATP was prepared from rabbit
muscle as the Ba salt by the method of D. M. Needham (194:2), and
as the neutral or acid Ba salt and the Ca salt by the method of Kerr
(1941). No Ba salt and only Ba-frec reagents were used in the pre-
paration of Ca ATP.

A solution of the Na ATP was ])reparcd from the Ba salt b}’’ addition
of the calculated amount of sodium sulphate to the Ba salt cither
suspended in water or dissolved in dilute hydrochloric acid. The Ca
ATP was converted into the Na salt by the addition of the calculated
amount of sodium oxalate to the solution of Ca ATP in dilute hydro-
chloric acid. Both the isolated Ba and Ca salts and the final Na ATP
solutions were analysed for N and for total and 7’ P according to the
method of Piske and Subbarow (1925) in the modification of Scheee
(1936). The purity of each substance was at least 98 — 99 per cent.

Inosine triphosphate (ITP). The substance Avas prepared as the Ba
salt from Ba ATP according to Kleinzeller (1942).

A solution of Na ITP was prepared by addition of the calculated
amount of sodium sulphate to the Ba salt suspended in water. Analysis
of both the Ba salt and the Na ITP for N and total and 7’P indicated
a purity of at least 98 — 99 per cent.

Creatine phosphate was prepared by the method of Zeile (1938)
by phosphorylation of creatine with POCI3. The isolated Ca salt was
first purified according to Zeile and then converted into the Ba salt,
which was recrystallized repeatedly from water-methanol according
to Deutsoh et al. (1938). Analysis: P, 8.4 per cent, N, 11.3 per cent;
CiHaOjNjPBa • HjO requires: P, 8.5 per cent, N, 11.5 per cent. A
solution of Na creatine phosphate was prepared by precipitation of
Ba with the calculated amount of sodium sulphate from the solution
of the Ba salt in water.

Acetyl phosphate. The substance was prepared according to Lynen
(1940) as the silver salt. (P, 8.0 per cent; calc, for CjHaOsPAgj’. P, 8.8 per
cent). A solution of sodium salt was prepared by addition of the calculated
amount of NaCl to the suspension of the silver acetyl phosphate in water.

Adenosine. The substance Avas prepared from yeast nucleic acid by
the method of Bredereck (1938) and recrystallized repeatedly from
Avater. (N, 26.2 per cent; calc, for OJ0II13O4N6: N, 26.2 per cent).

All substances Avere applied iso-osinotically by replacing part of the
NaCl 4- water in the Einger solution by a corresponding amount of
the substance dissolved in Avater, the solution being adjusted to pH 7.3.

Results.

1. Adonosine triphosphnto.

ATP applied to striated muscle showed the same effect with
respect to release of contraction and changes in birefringence
as described in our previous paper (1944), regardless of the method

328 I’KITZ BUCHTHAL, ADAM DEDTSCH AXD G. G. K^"API>EIS.

of preparation of the Ba salt or of the resulting Na ATP solution.
Since solutions containing Ba ions were very active when applied
to heart muscle (unpublished experiments of Deutsch andLuxmx),
we thought it necessary to investigate the effect of (1) Ba salts
on skeletal muscle, and (2) an ATP solution prepared with
complete exclusion of Ba ions during all phases of the prepara-
tion.

BaCL in the concentrations investigated (0.01 — 0.002 M) has
no effect on striated muscle fibres. Na ATP from the Ca salt
prepared according to Kekr, but without the use of barium salts,
acts like all other ATP preparations on contraction and birefring-
ence with a threshold value at a concentration of 0.72 x 10““
mol/ml.

In work on myosin, calcium and magnesium ions have been
shown to play an important part in the interaction between
myosin and ATP. In muscle the threshold for ATP increases
considerably with increasing concentration of calcium ions applied
isotonically. The threshold for ATP is about 50 times higher in
the absence of Ca or if it is present in excess (0.009 M CaClj*

= 0.1), corresponding to 5 times the Ca concentration
in normal Bingef solution.

While there is an optimum concentration for Ca, the presence
of magnesium ions increases the threshold value for ATP and for
electrical stimulation. With 0.01 M MgClj applied iso-osmo-
tically in Ringer solution the threshold is 10 times normal, while
it is still higher with 0.05 M MgCL.

2. Iiiosine ti’ipliospliate.

Inosine triphosphate is more active than ATP in releasing
contraction, the threshold concentration being 0.25 x 10 “
mol/ml. Larger amounts initiate repetitive acti^>ity of long dura-
tion. Birefringence remains constant even when the substance
is applied in a thirty times higher concentration than its threshold
for release of contraction.

3. Creatino phosphate.

It was of special interest to compare creatine phosphate and
ATP in their effects on the isolated muscle fibre, since the break-

Kl'PKCT OF ADKXO^INK TU!IMIO?I>ltATK.

329

down of tlioso two substances occurs nearest in time to the con-
traction process. Creatine phosphate neither releases contrac-
tion nor affects bircfrijigcncc (concentration applied 7.2 x 10“®
niol/nil).

"When applied Fiinultancously with small amonnt-s of ATP,
creatine phosphate increases the effect of the former, and can
release contraction anew in a muscle fibre in wliich contractions
evoked by ATP have aircad}' ccn.sc-d. This effect is ])robably due
to the regeneration of A'J']^ by the added crcjitine phosphate.

Pig. 2. liircfringcnrc ns fimofion of time after nj'plicntion of creatine pliospimfo
-f ndcnylic ncid (7.2 X 10”‘ mol/inl). Subslnnce.s npplied at 0 min.

Creatine phosphate + adcni/lic flctW- releases contraction (thresh-
old 3.G X 10”' inol/ml) and causes a reversible fall in bire-
fringence of about 15 per cent, in concentrations of 7.2 x 10””
mol/inl. But whereas with ATP birefringence attains its minimum
after 2 minutes and is completely restituted within 20 minutes,
after application of creatine phosphate -b adenylic acid restitution
begins after 12 — -15 minutes and is comjdetc onlj' after about
30 minutes (Ifig. 2).

4. Aoctylcholliio-adoiiosliie trJi)ho8i)lintc.

According to Feldbkrg and j\Iakn (1915) ATP is necessary for
the resynthesis of acetylcholine in cell-free brain extracts. "We
have sought for evidence for such an interaction in frogs muscle,
and investigated whether the local distribution of ATP in muscle
points to a close relationship of tliis substance with the function
of the motor end plate. The pelvic part of the frog’s sartorius
muscle (l/5th of its length) has been shown to bo entirely free

330 FRITZ BUCHTHAL, ADAM DEUTSCH AI?D G. G. KKAPPEIS.

from nervous elements (Metenay and Nachmansohn (1938),
P^ZARD and May (1937)). Phosphate analysis of this part of the
muscle reveals, however, no differences in ATP content in com-
parison with the rest of the muscle. In search for a common link
between acetylcholine and ATP we have examined another high-
energy phosphate, acetyl phosphate, the presence of which has
recently been demonstrated in living tissues by Lipmann 11912).
This substance, however, even in high concentrations (up to
14.4 X 10”*’ mol/ml) has no effect whatever on the contractility
or birefringence of non-curarized or curarized muscle fibres.

Fig. 3. Birefringence as function of time.

X — X — X - X application of ATP (7.2 X 10"' mol/ml)

• • application of ATP (7.2x10-'' mol/ml) after previous treatment

■with acetylcholine (10 /«g/ml). Substances applied at 0 min.

Acetylcholine, on the other hand, can interfere with the action
of ATP in striated muscle. It has previously been observed that
ATP sensitizes amphibian and mammalian muscle for subsequent
application of acetylcholine (Bucuthal and Kahlson (1944),
Buchthal and Folkow (1944)). This sensitization is, however,
localized to the contractile substance, as the response of curarized
muscle to direct electrical stimulation is likewise considerably
increased after previous application of ATP. When a muscle fibre
is treated "with minute amounts of acetylcholine (10 / 2 g/ml), this
substance eliminates the change in birefringence otherwise
present after application of ATP (Fig. 3). In denervated amphibian
muscle, jmst as in mammalian muscle (Buchthal and Kahlson
(1946)), previous treatment with acetylcholine ev'en abolishes
release of contraction by ATP.

EFFECT OF ADENOSINE TllIPIIOSPHATE.

331

Denervaied amphibian muscle Las a reduced sensitivity towards
ATP. A concentration producing a long-lasting repetitive activity
in curarized and non-denervated muscle fibres, evokes only a short
tetanic contraction in denervated muscle. Birefringence here
remains constant after application of ATP.

5. Adenosine.

Adenosine, like adenylic acid, has no effect on striated muscle
(concentration examined 7.2 — 15 x 10"® mol/ml), but applied
together with pyrophosphate it lowers the treshold of the latter
and at the same time changes the type of contraction from con-
tracture-like to repetitive tmtch-like activity. While pyrophos-
phate causes irreversible changes in birefringence the applica-
tion of adenosine -f- pyrophosphate leaves the latter unchanged.
It may be worth mentioning that aneurin + pyrophosphate applied
to the muscle have the same effect as adenosine -f pyrophosphate,
while aneurin without pyrophosphate has no effect whatsoever.
It is tempting to assume that the aneurin -f pyrophosphate effect
is due to the formation of cocarboxylase.

Discussion.

The action of adenosine triphosphate on striated muscle fibres,
as demonstrated in a previous communication, is twofold, viz.
it releases contraction and initiates reversible changes in bire-
fringence, In view of the time relations between mechanical
processes and changes in birefringence, the latter are considered
as an expression of resticutional processes in the contractile
proteins. The results are interpreted by assuming ATP to be the
normal agent of contraction, initiating a discharge of contractile
proteins, and the primary link in the transfer of energy for re-
stitution, i. e. recharge of contractile proteins. The effect is highly
specific for ATP and only adenosine diphosphate and adenylic
acid -}- pyrophosphate have the same mode of action. The latter
substances, however, can be considered as ATP precursors.

Inosine triphosphate is more active in releasing contraction
than ATP but is without effect on birefringence. Contraction
can also be evoked by inorganic ti’iphosphate, indicating that
the triphosphate part of the ATP molecule is sufficient for its

332 FRITZ BUCHTHAIi, ABAM DEUTSCH AI^D G. G. KXAPPEIS.

release.! On tlie otlier hand, the changes in birefringence caused
by ATP are highly specific.

This implies that restitution does not occur after appheation
of triphosphate or inosine triphosphate, although the latter only
differs from ATP in that an NH; group is replaced by OH. In
fact it is found that triphosphate only acts once when administered
by micro-application, while the effect of ATP can be repeated
several times.

The investigation of the effect of creatine phosphate was of
special interest, since the present experiments allow a direct
comparison of the two energy-supplying compounds, the break-
down of which occurs nearest in time to the contraction proper.
Although experiments on muscle extract favour the view that
the breakdown of ATP is the primary reaction, other investigators
(Lipjiann (1941)) consider the question still open, as processes
occurring in unorganised systems such as muscle extract cannot
be compared vith energy transfer in the intact muscle. The com-
pletely negative results with creatine phosphate regarding both
release of contraction and changes in birefringence, decide in
favour of ATP as the agent of contraction and the immediate
source of energy for the recharge of the contractile protein. ATP
can be substituted in its effect on muscle by creatine phosphate -f
adenylic acid, a fact which is in agreement with results from
work on muscle extracts, viz. that the hydroljttic breakdown
of creatine phosphate is coupled to the intermediate formation
of ATP.

Summarj'.

1. In agreement with previous results it was found that ATP
prepared in different ways initiates contraction and changes
birefrigence when applied in minute amounts to isolated striated
muscle fibres of the frog. ATP is evenly distributed in the
muscle and not localized specially in the region of the motor
end plate.

2. Lack or excess of calcium just as excess of magnesium
increase the threshold value for ATP.

! We ■vvisli to correct an erroneous statement in the discussion of our previous
paper (1D44, p. 2S5). Evidence was presented tliat the triphosphate part of the
ATP molecule was energetically and structurally sufficient to release contri-
tion, and not; as stated, both the triphosphate and nucleotide parts of the
molecule.

EFFECT OF ADENOSIKE TUIPIIOSPHATE.

333

3. While adenosine alone is irithout effect, it changes the con-
tracture-like effect of pyrophospate to a normal tetanus-like
contraction.

4. Inosine triphosphate, thrice as active as ATP in initiating
contraction is -without effect on birefringence.

5. Creatine phosphate has no effect on striated muscle, thus
establishing the breakdown of ATP as the reaction nearest in
time to contraction. Creatine phosphate -f adenylic acid can
replace ATP in the release of contraction and causes a reversible
fall in birefringence. The time of restitution, however, is consid-
erably longer than -with ATP.

6. Acetyl phosphate is -ndthout effect.

7. Acetylcholine abolishes the effect of ATP on birefringence
while it does not affect release of contraction. In denervated
muscle previous application of acetylcholine prevents release
of contraction by ATP. Contraction in denervated muscle is not
accompanied by the slow changes in birefringence.

The present work has been supported by grants from the
Michaelsen Foundation and by the Nordish Insulin Foundation,

Added in the proof : A paper by Toeda and Wolff (1946) cama
recently to our knowledge in which it is discussed, how far a
release of contraction by ATP may be due to the removal of
intracellular calcium by the liberation of phosphate. In this
connection it can be of interest to mention our experiments
on the effect of sodium citrate. The threshold concentration of
citrate in releasing contraction is higher than that of ATP.

Previous treatment of a muscle Avith strong concentrations of
citrate (3 x 10“® mol/ml) does not abolish the sensitivity to ATP.
No decrease in birefringence occurs after application of citrate
(3.6 X 10~“ mol/ml), while it inhibits the decrease in birefringence
caused by ATP. These experiments give no support to the as-
sumption that the action of ATP is due to the removal of intra-
muscular calcium.

Eeferences;

Bredereck, H., a. Martini, and F. Richter, Ber. dtsch. chem. Ges.
1941. 74. 694.

Buchthal, F., a. Deutsch, and G. G. Knappeis, Nature 1944. 153. 774.
' > — > — , Acta Physiol. Scand. 1944. 8. 271.

334 FRITZ BUCHTHAl, ADAM DEUTSCH AND G. G. KXAPFEIS.

Buchthal, F. and B. Fodkow, Ibidem 1944. 8. 312.

Buchthal, F. and 6. IIahlson, Ibidem 1944. 8. 317.

— , — , Ibidem 1946, in press.

Buchthal, F., and G. G. ELnappeis, Skand. Arcb. Physiol. 1938. 78. 97.
Dainty, M., A. Kleinzeller, A. S. C. Lawrence, M. Miall, J.
Needhasi, M. D. Keedham, and Shih-Chang Shen, J. general
Physiol. 1944. 27. 355.

Deutsch, a., M. G. Eggleton, and P. Eggleton, Biochem. J. 1938.
32. 703.

Feldberg, W., and T. Mann, J. Physiol. 1944. 103. 28 P.

Fiske, H. C., and Y. Subbarow, J. Biol. Chem. 1925. 66. 375.
Gronwall, a., and B. Ingeljian, Nature 1945. 155. 45.
Kleinzeller, A., Biochem. J. 1942. 36. 729.

Kerr, S. E., J. Biol. Chem. 1941. 139. 131.

Lipjunn, F., Advances in Enzymology 1941. 1. 149.

— , Federation Proc. 1942. 1. 122.

Lynen, F., Ber. dtsch. chem. Ges. 1940. 73. 367,

Mdrnay, a., and D. Nachmansohn, J. Physiol. 1938. 92. 39.
Needham, D. M., Biochem. J. 1942. 36. 114.

Pezard, a., and E. M. May, C. E. Soc. Biol, Paris 1937, 124. 1081.
ScHEEL, K. C., Z. analyt. Chem. 1936. 105. 256.

Torda, C. and H. G. Wolff, Amer. J. Physiol. 1946. 145. 419.

Zeile, K. and G. Fawaz, Hoppe-Seyl. Z. 1938. 256. 193,

Rate of Renewal of RiRo- and Dosoxyribo
Nucleic Acids.

E. HA7>IMARSTEK nnd G. HEVESY.

Rwfivrd 2 March 1916.

Enzymic proccssics coupled with phosphorylation take often
place at a remarkedl)’ rapid rate. A large percentage of the raole-
culcB of many of the acid-soluble phosphorus compounds and,
to a minor extent, also those of phosphatidcs present in the liver
and some other organ.s arc renewed within a short time. This is
demonstrated by the observation that shortly after the adminis-
tration of ’-P these molecules are found to contain labelled
phosphate.

That the presence of ’-P in the molcctjles of organic phosphorus
compounds indicates an enzymic sjmthcsis of such molecules
is most strikingly demonstrated by u recent experiment of Chat-
Korr and his associates (1912). These authors have shown that
labelled phosphatidcs arc formed when surviving liver slices
are shaken with bicarbonate Itinger solution containing labelled
phosphate, that this formation is, liowcvcr, impaired in the ab-
sence of oxygen, and homogenized liver tissue completely loses
its ability to incorporate ^*P into the phosphatide molecule. A
non-enzymic process could hardly be dependent on the intact-
ness of the tissue cells.

In contradistinction to the above mentioned compounds, as
found in previous (Hevesy and Ottesen 1943, AhlstbOm,
Euler and Hevesy 1944, Brues, Tracy and Cohn 1944) and
m the present investigations, dcsoxyribo nucleic acid molecules
present in the liver are renewed at a very slow rate only. This
result falls in line with the view that dcsoxyribo nucleic acid

336

E. HAMMAESTEN AKD G. HEVESY.

is present in the nuclei of the cells and is involved in the process
of cell division. As the mitotic process in the liver of a fully
grown rat takes place at a very slow rate only, the low rate of
renewal of desoxyribo nucleic acid in the fully grown liver is in
no way surprising, neither is so the much higher rate of renewal
observed in the liver of the only few days old rat. The rate of
formation of new desoxyribo nucleic acid molecules present in
the liver of 3 to 4 days old rats was found to be about 20 times
that of the corresponding figure in outgrown rats (Ahlstrom,
Euler and Hevesy 1944).

In contradistinction to the desoxyribo nucleic acid, a large
part of ribo nucleic acid is located in the cytoplasm and, according
to the view developed by Caspersson (1940), is involved in pro-
tein synthesis. As such a synthesis takes place at a marked rate
in the liver, the rate of renewal of the ribo nucleic acid can be
expected to be larger than the rate of renewal of desoxyribo
nucleic acid in this organ. The present note contains the results
of experiments in which the rate of renewal of both types of nucleic
acid was determined, vis. in the liver, the spleen, and the intestinal
mucosa of the rat and also in the total rat body. The methods
applied in the isolation of the nucleic acids will shortly be published
by one of the authors.

Kesults.

If we assume the free phosphate to be built into the nucleic
acid molecules without the formation of an intermediary phos-
phorus compound of comparatively long life, the ratio of the
specific activity of the nuclei acid P and the free P is a measure
of the rate of formation of new nucleic acid molecules and thus
of the extent of renewal of such molecules. If, previous to the
formation of labelled nucleic acid molecules, labelled precursors
would be built up at a rate slower than the rate of formation
of new nucleic acid molecules, the ratio of the specific activity
of the nucleic acid P and that of the free P would no longer be a
proper measure of the rate of renewal of the nucleic acid. In the
latter case, namely, during part of or possibly throughout the
whole experiment new nucleic acid molecules would be built up
without participation of ®^P. The participation of such labelled
intermediary compounds of a very long life in the formation
of desoxyribo nucleic acid in the liver is, however, highly improb-
able in view of the comparatively short lifetime of most of the

KIBO- AND DESOXYRIBO NUCLEIC ACIDS. 337

acid-soluble phosphorus compounds and the very long lifetime
of desoxyribo nucleic acid molecules present in the liver.

In our experiments the specific activity of the nucleic acid
P is compared T?ith the specific activity of the free P at the end
of the experiment. As the specific activity of the free P changes
throughout the experiment, the specific activity of the nucleic
acid P at the end of the experiment should, however, be compared
with the mean specific activity of the free P during the experiment.
As the liver takes up ®=P at a very rapid rate, and its free ®-P
content culminates within the first 2 hours, the end value and the
average value of the specific activity of the free liver P do not
differ essentially, the average value being about 6 per cent lower
than the end value (Ahlstrom, Euler and Hevesy 1944).
In the case of the spleen the corresponding figure is about 25,
and even larger differences are found in the case of the intestinal
mucosa. The figures of columns 2 and 3 of Table 1 should there-
fore be multiplied with 1.05 in. the case of the liver, for example,
to obtain a correct value for the amount of the rate of renewal
of the desoxyribo nucleic acid present in the Hver. In the figures of
Table 1 we have not introduced this correction, as we are mainly
interested in the relative rate of renewal of the desoxyribo and
ribo nucleic acids.

A further point which was not considered is the repeated
renewal of the same molecule during the experiment. In view
of the percentage renewal amounting in our experiment to few
per cent of less only, the probability of repeated renewal of a
molecule during the experiment is small, though the problem
whether all nucleic acid molecules present in the liver have the
same chance of being renewed or some of them are situated in
preferential districts of the cell and are thus renewed at a more
rapid rate is yet unsolved.

Table I contains the results of an experiment in which three
rats in nitrogen-equilibrium weighing 252, 182 and 215 g were
injected subcutaneously with respectively 7.5, 6.6 and 5.6 micro-
curies of 32P per 100 g body weight.

After 2 hours the animals were killed and the organs prepared
according to a method which will soon be published by one'of
the authors.

As recorded in Table 1, the rate of renewal of ribo nucleic
acid in the liver is as much as 33 times larger than the rate of
renewal of dexoxyribose nucleic acid. In spite of the finding that

22 ff60215. Acta phys. Scandinav. Vol. 11.

3S8 E. HAMMARSTEN ARD G. HEVESY.

ribo nucleic acid is renewed at an even larger rate in the spleen
and the intestinal mucosa than in the liver, the ratio of the rate
of renewal of ribo and desoxyribo nucleic acids in these organs
is only 3 and 2, respectively. This low ratio is due to the compara-
tively high rate of renewal of desoxyribo nucleic acid in these
organs. I?rom the above figures it follows that the rate of renewal
of both types of nucleic acid is highest in the intestinal mucosa
and in the spleen.

Table I.

Batio of the rate of renewal of the ribo nucleic acid and desoxyribo
nucleic acid in the organs of the rat in the course of 2 hours.

Organ

Percentage ratio of the
specific activity of the
nucleic acid P and
that of the free P

Ratio of the
rate of
renewal of
ribo and
desoxyribo
nucleic acid

Ribo

Desoxyribo

nucleic acid

Liver

3.8; 3.6

0.12; 0.09

Spleen

3.1; 10.2

2.2: 2.2

Intestine

7.1; 4.1

3.4; 2.3

The specific activity of both the desoxyribo and the ribo
nucleic acid P extracted from the rat liver was determined by
Brtjes, Tracy and Cohn (1944) in experiments lasting 3 to 8
days. In these experiments the ribo nucleic acid P was found to
be only 5 to 6 times as active as the desoxyribose nucleic acid P .
The discrepancy between these figures and those obtained by us
may be due, at least in part, to the much longer duration of
the experiment.

Specific Activity of the Nucleic Acid Phosphorus
Extracted from the Total Bat.

In another experiment the. specific activity of both the total
desoxyribo nucleic acid P and the total ribo nucleic acid P ex-
tracted from a rat weighing 194 g was determined. The activity
of labelled sodium phosphate amounted to 8.1 microcuries per
100 g animal weight. The time of the experiment was 2 hours.
The results of this experiment are seen in Table 2.

EIBO- AKD DBSOXYRIBO NUCIiEIC ACIDS.

339

Tabic 2.

Specific activity of the nvcleic acid P extracted from the total rat
compared with the corresponding values for liver, spleen, and
intestinal mucosa. The value for the specific activity
of the total rat ribo P is assumed to be — 100.

Sample

Specific activity

Ribose

Desoxyribose

Free P

nnclcic acid

Total rat

100

Liver

164

4.4

5100

Spleen

292

2850

Intestine

112

2770

As shown in Table 2, the specific activity of the average nucleic
acid P of the rat is almost identical with the corresponding value
of the ribo and desoxyxibo nucleic acid, respectively extracted
from the intestine.

^ The interpretation of the significance of the specific activity
figures obtained for the total rat encounters some difficulties, as
the specific activity of the free P utilized in the formation of the
labelled nucleic acid molecules is unknown. If the specific activity
of the free P utibzed in building up the average body nucleic
acid would correspond to the specific activity of the free liver P,
the percentage "rate of renewal” of the body ribo and desoxyxibo
nucleic acids would be 2.0 and 1.2, respectively. If the specific
activity of the free P utilized in building up the average nucleic
acid of the organism would correspond to the specific activity
of the free intestinal P, larger figures, i. e. 3.6 and 2.2, respectively,
would be obtained.

When calculating the last mentioned figures we compared
the specific activity of the nucleic acid P at the end of the ex-
periment with the specific activity of the free intestinal P at
the end of the experiment. Correctly we should have considered
the mean value of the specific activity of the three intestinal
I* prevailing during the experiment. The mean value of this
magnitude is about almost half of its end value, we have there-
fore to multiply the figures mentioned above (3.6 and 2.2 respec-
tively) with about 2 to obtain an approximate value of the per-

340

E. HAMMABSTEN AND G. HEVEST-

centage renewal of tlie ribo- resp. desoxyribo nucleic acid in the
course of 2 hours.

It is improbable that a so highly active free phosphate is
utilized in the building up of the nucleic acid molecules as found
in a 2 hours experiment in the liver. Liver and kidneys have a
privileged position concerning the rate of intrusion of phosphate.
The amount of nucleic acid present in the liver and the kidneys
makes out, furthermore, only a small percentage of the total
nucleic acid content of the organism. It is much more probable
that free P of similar specific activity as found in the intestine
is applied in the building up of the labelled nucleic acid molecules.
In fact, the amount of nucleic acid present in the mucosa of
the digestive tract makes out a large percentage of the body
nucleic acid. While the body nucleic acid contains also slightly
radioactive fractions, viz. those originating from the liver, the
kidneys, and the brain and fractions of restricted radioactivity
originating from the muscles (Hevesy and Ottesen 1943),
it contains also fractions of higher activity than found in the
intestinal mucosa, viz. those originating from the bone marrow,
the thymus and lymph nodes (Andbeasen and Ottesen 1944,
1945). The lymphocytes secreted into the organism can also
be expected to contain pronouncedly active nucleic acid. This
makes it understandable that the rate of renewal of the average
body nucleic acid corresponds to about the rate of renewal of the
intestinal nucleic acid and is thus quite pronounced for both types
of nucleic acid in contradistinction to the rate of renewal found
in the liver, which is very low in the case of desoxyribo nucleic
acid and appreciably higher in the case of ribose nucleic acid.

Discussion.

The mitotic figure of the fully grown liver is unknown. In
hemectomized livers of quite young rats (50 g) figures of 1 — 2
per cent are recorded (Mabshak and Walkeb 1945). Even if
the formation of labelled desoxyribo nucleic acid molecules in
the fully grown rat is in direct connection with mitotic pro-
cesses, the mitotic figure can not be as high as 0.1 per cent
(a mitotic division lasts about 1 — 2 hours), since in some phases
of the mitotic process nucleic acid disappears, in others accu-
mulates, as demonstrated by Caspebsson (1940). When com-
paring, furthermore, the number of labelled desoxyribo nucleic

MBO- AND DESOXYRIBO NDCbEIC ACIDS.

341

acid molecules formed in growing Jensen sarcomata of the rat
with the increase in the desox 3 Tribo nucleic acid content due to
growth, the former figure was found to be higher than the latter
(Euler and Hbvesy 1944) and a similar result was obtained
when comparing the formation of labelled desoxyxibo nucleic
acid molecules in 3 days old rats with the increase in the desoxy-
ribo nucleic acid content due to growth of the liver.

In view of the high desoxyribo nucleic acid content of the
lymphocytes and because they arc partly produced in the spleen,
the comparatively high rate of turnover of desoxyribo nucleic
acid in the spleen is in agreement with our expectance.

It is more difficult to interpret the high figures found for
the rate of renewal of desoxyribo nucleic acid in the intestinal
mucosa. The intestinal mucosa is exposed to very hard tear, cells
are destroyed at a large scale and new ones are formed. It is,
however, very problematic whether this process alone can account
for a so high rate of renewal of the desoxyribo nucleic acid as
found in this organ.

If we accept the view put forward by Caspersson, the high
rate of renewal of ribo nucleic acid is in no way surprising. That
the high figures found for the rate of renewal of ribo nucleic
acid in the intestine, the spleen and the liver is just what we
would expect in view of the importance of these organs in protein
metabolism. The incorporation of labelled sulfur into protein
sulfur is found to be higher in the intestine than in any other
organ (Tarver and Schmidt 1942) and the content of the
proteins isolated from the intestinal wall of the rat after adminstra-
tion of isotopic 1( — ^)-leucine is larger than the corresponding
value for any other organ investigated. Somewhat smaller
values for the content of the proteins isolated from the spleen
were found, and still smaller values for the of the proteins
isolated from the liver (Schoenheimee, Eatner and Bittenbeeg
1939). The rate, of formation of rihose nucleic acid in these three
organs diminishes in the same sequence.

If we want to state, not as above the percentage, but the
amount of nucleic acid formed during the experiment, we mush
know the nucleic acid content of the organs of the rat and of the
total rat.

Some preliminary figures for the total nucleotide P of the liver,
spleen, intestine and total rat and also some preliminary figures
or the share fo polydesose and polyribo nucleotides in tbe total

342

E. HAMMAKSTEN AKD G. HEVESr.

nucleotides is seen in Table 3. Tbe method applied in obtaining
these figures and more accurate data will be shortlj^ published
by one of the authors.

Table S.

Pohjdesose nucleotide phosphortis and Folyriho nucleotide phosphorus
content of some organs and of the total rat.

Approximate
share of
polydesose
nnelcotides
in the total
nncleotides

g nucleotide
P per 100 g
dry weight

g polydesose
nucleotide

P per 100 g
dry weight

g polyribo
nucleotide

P per 100 g
dry weight

Total rat

45—50

0.232

0.11

0.12

Liver

35 %

0.350

0.12

0.23

Spleen

75 S

0.643

0.48

O.IG

Intestine

57 f.

0.6G9

0.38

0.29

Assuming the percentage renewal of the polydesose nucleic
acid of the total rat in the course of 2 hours to be 4 (cf.p,339)
and the fresh weight of the rat to amount to five times its dry
weight, in a 200 g rat in the course of 2 hours about 2 mg polydesose
nucletoide P will be renewed. The corresponding figure for the
polyribo nucletoide P works out to be 3, In the total rat the
turnover rate of the 2 types of polynucleotides does thus not
differ very appreciably.

A very different result is obtained when comparing the amount
of polydesose and polyribo nucleotide phosphorus renewed in the
liver. The figures work out, assuming the liver to weigh 6 g,
to be 0.0017 mg and 0.094 mg respectively. Fiftyfive times more
polydesose nucleotide than ribo-nucleotide gets thus renewed in
the liver during the same time.

Assuming the spleen to weigh 0.8 g, both the amount of poly-
desose nucleotide P and that of polydesose nucleotide P formed
and still present in the spleen works out to be about 20 microgram.

How far the rate of ensymic replacement of other constituents
of the nucleic acid molecule, for example that of the pyridine
and pyrimidine groups, takes place at a similar rate as the en-
zymic replacement of phosphorus is not yet elucidated.

MBO- AND DESOXYRIBO NUCLEIC ACIDS.

343

Summary.

Labelled sodium pbospliate is administered to rats and after
tbe lapse of 2 hours the specific activity of tlie ribo-nucleic acid
phosphorus and that of the dcsoxyribo-nucleic acid phosphorus
determined.

In the liver the specific activity of the ribo-nucleic acid P is
found to be 33 times larger than the specific activity of the
desoxyribo-nucleic acid P. In the course of 2 hours about 0.1
and 3.3 per cent respectively of these compounds were found
to be renewed.

In the intestine and in the spleen in which the specific activity
of the desoxyribo-nucleic acid is found to be about 20 times
larger than the corresponding value in the liver, the specific
activity of the ribo-nucleic acid phosphorus is only 2 to 3 times
larger than the corresponding value of the desox}T:ibo-nucleic
acid phosphorus.

The ribo- and the desox)rribo-nucleic acid phosphorus extracted
from the total rat have a very similar specific activity to the
corresponding phosphorus extracted from the intestine. In the
total rat the difference in the rate of renewal of the two types
of nucleic acid is not very pronounced. In a rat weighing 200 g
approximately about 2 mg desoxyribo-phosphorus and 3 mg ribo-
nucleic acid phosphorus are turned over in the course of 2 hours.

Beferences.

Ahlstrom, It,, H. V. Euler, and G. Hevesy, Sv. Vet. Akad. Ark.
Kemi 1944. 19 A. no. 9.

Andreasen, E., and J. Ottesen, Acta path, microbiol. scand. 1944.
suppl. 54.

Andreasen, E., and J. Ottesen, Acta physiol, scand. 1945. 10. 258.
Brues, a. M., M. Tracy, and W. E. Cohn, J. biol. Chem. 1944. 155.
619.

Caspersson, T., Chromosoma 1940. 1. 662.

Chaikopf, I. L., Physiol. Rev. 1942. 22. 291.

V. Euler, H., and G. Hevesy, Sv. Vet. Akad. Ark. Kemi 1944. 17 A.
no. 30.

Hevesy, G., and J. Ottesen, Acta physiol, scand. 1943. 5. 237.
Marshak, A., and A. C. Walker, Amer. J. Physiol. 1945. 143. 251.
ScHOENHEBiER, R., S. Ratner, and D. Rittenberg, j. biol. Chem.
1939. 130. 703.

Tarver, H., and C. L. A. Sohmidt, Ibidem 1942. 146. 69.

From the Department of Medical Chemistry; Dniyersity of Uppsala.

On the Pui’ification of the Thiiimiii-InactiTatiiig
Pish Pactor II.

GUNNAB, AGREN.

Received 13 March 1946.

A method was recently described for the tenfold purification
of the thiamin-inactivating fish factor (Agben 1945 a). By means
of this method definite proofs were obtained that the thiamin-
splitting principle was not a haemin protein. On the other hand,
the purified solution of the factor was always yellow-coloured
and it had to be settled whether this circumstance was a mere
coincidence. The present paper reports the further purification
of the factor to a colour-free substance.

Experimental.

The enzyme was prepared from the viscera of ide according to the
above mentioned method (Agben 1945 a) and subsequently stored
at — 15° C. The viscera were collected under the same precautions
as previously described (Lieck and Agben 1944). The activity of the
factor was ascertained by a method based on the diazo reaction of
Pbeblttda and McCollum. This method was a modification of the
procedure described by Melnick and Field 1939 (cf. Lieck and Agben
1944). The activity was expressed in units as previously defined (Agben
1945 a).

Results.

The activities of the solutions of the factor purified according
to the above-mentioned method were rather unstable, even
when the material was stored at — 15° C. Possible means of

THUMIN-INACTIVATING FISH FACTOR.

345

further purification seemed, ho-wever, to he few. Knally a separa-
tion of impurities was attained by means of cataphoresis. As
previously stated (Agren 1945 a), a preliminary cataphoretic
analysis of the purified solutions of the factor usually revealed
the presence of three fractions, all of which seemed to move
parallel at different pH values. A further analysis now showed
that a separation could be obtained by cataphoresis at pH 4.7.

Fig. 1. Electrophoretic diagram obtained from an extract of the purified factor.
For experimental conditions see the text.

Visual observations and photographic records obtained with
the optical arrangements, described by Svensson (1939), disclosed
the presence of two components of different electrochemical
behaviour (Fig. 1), The activity was associated with the small
fraction which exhibited a slow cathodic mobility, while a larger
fraction was rather stationary throughout. The results were
easily reproducible with extracts of ide (cf. discussion).

Solutions for cataphoretic purification were prepared on a large
scale in the following way. Samples of 1 — 2 kg of viscera (liver,
spleen, intestines and gills) from about 30 kg of ide were purified
as previously described (Agren 1945 a). The purified solutions
usually contained about 2 mg of protein nitrogen and 3 — 4 units
of activity per ml of solution. 40 ml of this solution was dialyzed
in a cellophane tube (0 = 27 mm) at 0° C. for 1 hour against
distilled water and for 2 hours against the cataphoretic buffer
on a cradle dialysis apparatus. The contents of the cellophane
tube where then centrifuged from a precipitate which formed
during the dialysis. The centrifugate was dialyzed for 1 hour
against distilled water under the same conditions as described
above and afterwards concentrated in vacuo at -j- 15° C to about

346

GUNNAB AgREN.

15 ml. The concentrated solution was dialyzed for 4 hours against
the cataphoretic buffer solution under the same conditions as
described above. The buffer solution was renewed after 60 and
150 minutes. The contents of the cellophane tube were centrifuged
from a precipitate which formed during the dialysis. The centri-
fugate contained about 2 mg of protein nitrogen per ml. The
ionic strength of the buffer was 0.15 (0.05 of acetate buffer
+ 0.1 of sodium chloride). The centrifuged solution was placed
in a Tiselius electrophoresis apparatus of the usual construction
with capillary levelling system, and run for 28,800 seconds at
30 mA. After this time the small fraction alone filled the top
cathodic cell and was collected separatedly.

The active fraction with cathodic migration was only slightly
yellow in colour. The main part of the colour remained in the
stationary fraction. The contents of the top cathodic cell contained
about 0.13 mg of protein nitrogen and about 3 units of activity
per ml of solution or, calculated per mg of nitrogen, about 20
units, which implies a tenfold additional purification of the factor.
It was planned to collect the contents of the top cathodic cell
from 15 cataphoretic experiments in order to carry out a new
cataphoresis with this material. The series of experiments was
made and the collected contents of the top cathodic cell were
prepared for a new cataphoresis in the following way. The slightly
yellow solution, with a volume of 65 ml, was dialyzed at 0° C
for 1 hour against distilled water as described above. A rather
heavy precipitate which formed in the cellophane tube during
the dialysis was centrifuged off. The centrifugate was still slightly
yellow-coloured. The precipitate was dissolved in 10 ml of dis-
tilled water by adjusting the hydrogen ion concentration to
pH 7.4. The solution was clear and colourless and contained
about 85 per cent of the original activity in the cellophane tube.
The activity calculated per mg of protein nitrogen was still about
20 units (22). This figure certainly did not represent the maximum
activity obtainable. The series of electrodialyses was carried
out during the course of about 2 weeks. On the basis of previous
experience it was to be suspected that the activity of at least
the first samples in the cataphoretic series was partially destroyed
during storage in the frozen state. The activity must also have
been diminished by the repeated dialysis, which slowly splits
the enzyme into two inactive components (Agren 1945 a andb).

Several series of cataphoretic experiments were carried out

THIAMIN-INACTn’’ATlNG PISH FACTOR.

347

in tins manner, but the results were not always reproducible. It
sometimes happened, especially when the activity of the cata-
phorized and stored material was low, that the active substance
was not precipitated in the subsequent dialysis against distilled
water. The solution of the active precipitate gave the usual colour
reactions for amino acids.

Discussion.

Trom a previous work (Lieck and Agren 1944) it was known
that extracts of carp viscera destroyed more thiamin than simi-
larly prepared extracts of viscera from ide. It was therefore
planned to purify the thiamin-splitting factor from carp viscera.
"When a comparison was made of the activities of water extracts
of viscera from carp and ide prepared in the same way, they
were usually found to contain about 0.6 and 0.15 units of activity
respectively, calculated per mg of protein nitrogen. However,
it was soon found impossible to apply the previously outlined
method of preparation (Agren 1945 a) to such extracts. Several
other methods generally used in the purification of protein com-
pounds were unsuccessful when tried on extracts of carp viscera
of different dilutions and pH values. At present extracts of ide
are preferred for purification of the thiamin-splitting factor.

The biological significance of the thiamin-destroying factor
is still unclear. In a previous paper it was suggested that the
factor might be engaged in the control of the synaptic trans-
mission of nerves (Agren 1945 b). This hypothesis would seem to
presume a wide spread distribution of the factor. At present it
has only been found in different tissues of fish. A second possibility
must also be considered. In a recent paper Belopp and Stern
(1945) demonstrated that yeast carboxylase of different states
and degrees of purity was appreciably inhibited by treatment
with extracts containing the thiamin-destroying fish factor.
The reaction probably involved a destruction of the cocarboxylase
component of the enzyme. The authors did not discuss the
physiological significance of their experiments. It is clear, however,
that if the results are confirmed with carboxylase preparations
from fish, the thiamin-inactivating factor may be involved in
the regulation of oxygen consumption in a manner similar to
that exhibited by vitamin E, which has an inhibitory effect

348

GUNNAK AGBEN.

on the oxygen consumption of the succinic dehydrogenase S 3 ratem
(HotrcHiN 1942). Since it has been demonstrated in the present
paper that the thiamin-inactivating fish factor appears to be a
colour-free protein, it vould be of interest to investigate •whether
the dialyzable co-factor is identical •vrith glutathione or some
other reducing substance (cf. Ageen 1945 a and b).

Summary.

By cataphoretic separation and dialysis the thiamin-destroying
fish factor has been further purified about 10 times, in all about
100 times. The activity seems to be associated "with a colourless
protein.

The author is indebted to the Astra Corporation for a grant
"which supported the present investigation. He further expresses
his thanks to hir. Bklunl for valuable assistance throughout the
investigation.

Eeferences.

Beloff, R. and K, Stern, J. Biol. Chem. 1945. 15S. 19.
Houchin, J. Ibidem 1942. 146. 313.

Lieck, H., and H. Field, This journal. 1944. 8. 203.
Melnick, D. and H. Field, J. Kol. Chem. 1939, 127. 515.
SvENSSON, H. Kolloid. Z. 1939. 87. 181.

Agren, 6 ., This journal. 1945 a. 9. 306.

Agren, G., Ibidem 1945 b. 9. 221.

From the Biochemical Institute, Aarhus University, Denmark.

The Gastric Lipase in Man.

FRITZ SOH0NHEYDER and KIRSTEN VOLQVARTZ.
Received 23 March 1946.

The literature on the gastric lipase has been thoroughly reviewed
by Oppenheimeb (1926, 1936) and since then very few papers
on the gastric lipase have appeared. Whereas Volhard and his
school were inclined to ascribe a considerable importance to the
gastric lipase in the digestion of fats in the stomach both in dog
and man, it is at present taught that in the adults the gastric
lipase is of little or no importance in the digestion of fat. As re-
gards the importance of the gastric lipase in infants the question
is less clear. According to Oppenheimeb the existence of a charac-
teristic gastric lipase has never been proved in the human gastric
juice, as it is claimed that in the investigations there has not
been paid sufficient attention to regurgitation of pancreatic lipase
from duodenum to the ventricle. As late as in 1928 Melli and
Radioi have published results which indicate that the lipase
present in gastric juice from normal persons most likely is identical
with the pancreatic lipase. The majority of investigations on the
gastric lipase are rather ancient and have been carried out mostly
with fats which were not well defined.

We have been able to confirm that gastric juice from adults
and infants contains a characteristic gastric lipase, which is
identical with the lipase present in the gastric mucous membrane.
It is shown that as in the case of pancreatic lipase the optimum
pH for gastric lipase is very dependent on the triglycerides used
as substrate, the optimum pH for the lower triglycerides being
about 6.5 and for the higher triglycerides about 7.5. In case
of the higher triglycerides CaCU is able to shift the optimum

350

FRITZ SCH0NHEYDER AND KIRSTEN YOLQVARTZ.

pH to the acid side. The human gastric lipase is a very stable
enzyme in the acid medium. In vivo only the lower triglycerides
seem to be hydrolyzed to an appreciable degree by means of the
gastric lipase and apparently there is no basis for assuming
that this enzyme plays a greater role in the fat digestion in infants
than in adults.

Methods.

Sitbstrales. The substrates employed were; tripropionin, tributyrin,
tricaproin, tricaprylin, tricaprin, trilaurin, tristearin, triolein, cow-
butter fat and woman’s milk fat.

Enzyme. As enzyme preparations were used either glycerol extracts
of dried gastric membrane or aspirated fasting gastric juice. Glycerol
extracts of the dried mucous membrane were prepared according
to directions given in Oppenheimer (1925) with slight modifications.
Very shortly after death the stomachs of clinically healthy men, killed
in accidents, were opened and the mucous membrane was carefully
rinsed with water and wiped. The mucous membrane of fundus ventri-
culi was then scraped off, comminuted and dried with acetone, acetone-
ether and ether. The dried powder was extracted with 1 /iO n NH4OH.
The extract was precipitated ivith acetic acid and the precipitate dis-
solved in water by addition of small amounts of NaOH to get a clear
solution. Glycerol was added and the solution was concentrated in
vacuo, 2 mm Hg, to a mixture containing about 50 % of glycerol.
In the experiments on the determination of the activity of the extract
towards different triglycerides extracts and dilutions Yco '^Fere used.
Gastric lipase in gastric juice from adults was obtained by aspirating
the total fasting secretion. In order to get gastric juice from children
it was necessary to rinse the stomach with a small amount of water,
as generally no fasting secretion could be aspirated without this measure.
The fasting secretions in adults were occasionally coloured by bile,
whereas the rinse-water from children was always colourless. As will
be seen later, there was in no case any reason to assume that these
secretions contained other lipases than the characteristic gastric lipase.
In order to preserve the lipase in the secretions these were brought
to a pH about 5.6 immediately after the withdrawal.

Determination of the Enzymatic Hydrolysis, a) Experiments in Vitro:
The lipolytic activities of solutions containing gastric lipase towards
different triglycerides were determined as in our previous papers,
Scn0NHEYDER and Yolqvartz (1944, 1945). The amount of acid libe-
rated during the reaction was neutralized by adding dropwise 0.1 n
NaOH, keeping pH almost constant, the temperature being 40° C
i 0.5° C. The buffers used were acetic acid-sodium acetate and 5Ii-
chaelis’ veronal buffer, both O.l n. Regardless of solubility 0.263
millimol triglyceride in 30 ml reaction mixture were used for each
experiment. The initial velocity' was calculated graphically and given

THE GASTBIC LIPASE IN MAN.

351

as number of drops of O.i n NaOH per 10 minutes (1 ml = about 45
drops). The accuracy of the determinations of the lipolytic activity
was about 5 %.

b) Experiments in Vivo: The lipolysis in the stomach was tested
in the following way. In the adult the fasting content of the stomach
was aspirated and the stomach rinsed twice or more with water until
the aspirated water was clear and the stomach emptied completely.
Then the following test meal was given. 7.5 g peptone and V 200 mole
of triglyceride in 150 ml of water. The meals containing the lower
triglycerides and triolein were easily emulsified, whereas emulsions
of the higher triglycerides were more difficult to obtain. They were
melted, peptone solution added and by careful stirring a fairly good
emulsion was produced. The meal was ingested through the stomach
tube. 25 minutes after the finished ingestion the total content of the
stomach was aspirated through the tube. This experimental period
was chosen because the pH of the contents of the stomach from that
time generally reached a value at which no significant lipolysis is
supposed to take place. Peptone was added as a buffer substance.
The small amount of triglyceride was chosen on account of the difficulty
of obtaining these substances in pure state. In children the test meal
was given after previous washing of the stomach with a few ml of
water, as the stomach generally was empty. The test meal for children
contained the same amount of triglyceride as in adults but only 1.5 g
peptone i 150 ml.

On the material recovered from the stomach determinations of total
and free acids were carried out. Control experiments with some of the
subjects examined showed that their fasting secretions contained no
free fatty acids. In the case of tributyrin meals the amount of free
acid was determined by extracting 5 ml of the stomach content with
4 times 10 ml petrolether. The aqueous phase contained all the free
acid and no triglyceride, whereas the petrolether contained all the
triglyceride. In case of tricaproin and tricaprylin the free acid was
somewhat soluble in petrolether and therefore the meal was neutralized
before extraction. The aqueous phase contained all the free acid as
salt. The amounts of free butyric, caproic and caprylic acid were de-
termined by steam distillation of the acid into a receiver containing a
known amount of hydrochloric acid. COa-free air was bubbled through
the distillate for 15 minutes before titration with ^lo — h NaOH.
The total amount of these acids were determined by saponification of
suitable amounts of the stomach content and subsequent distillation
of the fatty acids. In the case of the higher triglycerides the triglyceride
and the free acid were both extracted from the stomach content,
made strongly acid, by means of petrolether. The aqueous phase was
discarded. The petrolether phase was washed with water and dried
with sodium sulphate. The petrolether was then distilled off and the
fatty residue dried to constant weight. The residue was then dissolved
in alcohol-ether mixture and the free acid titrated with — ^/so

NaOH. The total amount of fatty acid was either calculated from the
weight of the residue or determined by saponification of the residue.

352

FRITZ SCH0NHEYDER AHD KIRSTEK VOLQVARTZ.

Both in the case of the lower and the higher triglycerides the tech-
nique has been checked on known solutions of free acids with and with-
out triglycerides.

In a few experiments the degree of hydrolysis of cow-butter fat
and woman’s milk fat was determined. The woman’s milk fat was
prepared by extracting -the upper fatty layer of the centrifuged millf
with petrolether. The petrolether phase was washed with water and
dried with sodium sulphate and the petrolether was then distilled off.^
The test meals contained fatty substance corresponding to about
equivalents of fatty acids. The degree of hydrolysis was in these
cases calculated as the difference between the percentage of equivalents
of free acids in the aspirated meal (both in the petrolether extract
and the aqueous phase) and in the fat before ingestion.

Experimental.

A. The Optimum pH of Human Gastric Lipase with and without
Addition of CaCl^ to the Enzyme-Substrate System. Davidsohn
(1912) and Haueowitz and Petrou (1925) and Gyotoktj (1928 a)
have determined the optimum pH of the human gastric lipase
in the gastric secretions using the stalagmometric method and
tributyrin as a substrate. They found pH optima about 5 to 6.
ScHONHEYDER and VoLQVARTZ (1945) have in experiments with
pancreatic lipase found, that with increasing number of carbon
atoms in the fatty acids of the triglycerides the optimum pH
is displaced towards higher values, and therefore similar experi-
ments might be of interest in connection with the gastric lipase.
Our experiments with glycerol extracts of dried gastric mu-
cous membrane are carried out with and without addition of
calcium chloride (60 mg per 30 ml reaction mixture). The results
of the experiments with tripropionin, tributyrin, tricaproin, tri-
caprin, trilaurin and tristearin are given Kg. 1. The ordinates
give the initial velocities in per cent of the maximum velocity
in the system containing CaCls. The maximum velocity was
calculated graphically. In the experiments without addition of
calcium chloride it is seen that the optimum pH for the lower
triglycerides (tripropionin, tributyrin, and tricaproin) is found
between 5.5 and 5.8. By increasing number of carbon atoms in
the fatty acids there is an evident displacement of the optimum
pH to the alkaline side, the optimum pH for tricaprin, trilaurin
and tristearin being 7.2, 7.3 and 7.9 respectively.

‘ Tho authors are indebted to the Chief of the Lying-in Hospital for .Tntland,
Axei, Oesex, M. D., for supplying the milk samples.

THE GASTRIC LIPASE IN MAN.

353

23 — i60215. Acta phys. Scandinav. Vol. 11.

354 FBITZ SCH0NHEyDER AKD KIRSTEN VOLQVARTZ.

The question concerning the influence of the calcium ion on the
activity of gastric lipase has apparently not been investigated
previously. Our experiments show that in all the lower triglycerid-
es investigated the addition of calcium chloride to the enzyme-
substrate system is found to inhibit the activity of the enzyme
at pH-values greater than optimum pH. In the case of trilaurin
and tristearin where the free fatty acids are insoluble in water
there is a definite activation of the process at pH lower than 7,
and the optimum pH for the process is shifted 1.5 to 2 pH units
to the acid side. In the case of these triglycerides a definite in-
hibition is stated at pH higher than 7. The triglyceride tricaprin
takes up an intermediate position. As triolein was only split in
a very low degree by our glycerol extracts both with and ivithout
addition of calcium chloride the experiments on this substrate
are not included in Kg. 1.

Is should be mentioned that the pH-activity curve for tributyrin
was determined in one case with a fasting secretion from an adult
person. In this experiment a curve was found with quite the same
shape and pH optimum as the curves in Fig. 1 (II). Apparently
the enzyme in fasting secretion is therefore identical with the
enzyme prepared by extraction of the dried mucous membrane.

B. The Eelaiive Activity of Gastric Lipase iotoards Different
Triglycerides. The experiments in Fig. 1 have all been carried
out by means of extracts of the same gastric mucous membrane.
The activities of the extracts used towards the different triglyce-
rides, calculated for the same amount of enzyme and number
of equivalents of substrate, were very different, which appears
from Table I. In this table are given the relative initial
velocities (Elrd) at the respective pH-optima. is given for
systems wth and without CaCl-, and the velocities are related
to an arbitrary value of 100 for tributyrin. It is seen that the rela-
tive activities towards the solid triglycerides are very small.

The absolute value of activity towards tributyrin of 1 ml
fasting secretion from adults is about 0.45 ml of i/io n NaOH
per 10 minutes in the reaction mixture previously described.

C. The Stability of Gastric Lipase at Different pH-vahies. Inve-
stigations on the stability of the lipase present in the gastric
secretion of man at different pH-values are not known. From
our experiments (see Fig. 2) it appears that the enzyme is very
stable in the pH range from 3 to 7 and that in the range between
2 and 7.5 the half decomposition time for the enzyme is more

THE GASTRIC LIPASE IN MAN.

355

Table I.

Relative Initial Velocity of Hydrolysis of Triglycerides by
Human Gastric Lipase.

Triglyceride

Krol.

without CaCIj

Krol.

with CaClj

Tripropionin

Tribntyrin

100

Tricaproin •

Tricaprin

Trilaurin

Tristearin

0.8

Triolein

negligible

than 7 hours, the fasting secretions being kept at 40° C. The
different hydrogen ion concentrations were obtained by adding
HCl and NaOH to the secretions. The inactivation follows the
equation of the monomolecular reaction, thus the half decompo-
sition time can be determined fairly accurately. It appears that
the gastric lipase in man is a very stable enzyme, and there is
hardly any doubt that considerable amounts of gastric lipase
get into the intestine in non-destroyed condition, where it may
support the pancretaic lipase in the digestion of fatty substances.
Also Gyotoku (1928 b) is of the opinion, that gastric lipase in
man is an enzyme that is remarkably stable in comparison to
other lipases, but his investigations are carried out on extracts
of gastric mucous membrane, and the pH values at which his
examination of the stability were carried out are not given.

Dstermination of the stability of human gastric lipase in gastric juice at
different pH at 40° C. The ordinates represent the half decomposition time in

minutes.

356 FKITZ SCHONHEYDER AND KIRSTEN VOLQVARTZ.

D, Experiments on the TApolysis in the Stomach. The total gastric
digestion of fat in human subjects can not be determined, as such
determinations necessitate a duodenostomy or duodenal fistula
from which the gastric discharge can be collected as it is being
evacuated through the pylorus. The only satisfactory alternative
is to withdraw the gastric content by means of a tube. This only
jdelds information on the state of digestion at the time the
samples were obtained, but when the withdrawal takes place
at the same time after the intake of the meal it is possible to get
a relative measure of the digestion of the different triglycerides in
the stomach, and one can state whether there is any quantitative
difference between the digestion of fat in adults and in children.
The technique has been reported above, and 6 adults and 6
children were used as subjects for these experiments. The adults
were 3 males and 3 women, clinically normal, only No. 6 was
suffering from achyh'a. The 6 children were younger than 9
months and they were patients in the Department of Pediatrics,
Aarhus Kommunehospital.^ The experiments took place at the
time at which the children were convalescent after diseases,
which could not be supposed to influence the stomach function.
The results obtained from 18 test meals in children and 20 test
meals in adults are summarized in Table II. For each aspirated
meal the pH and the percentage of fatty acid liberated in relation
to total amount of fatty acid in the test meal are given. A correc-
tion for free acid in the triglyceride ingested has of course been
introduced. The degree of hydrolysis relates only to the state of
lipolysis of the triglyceride remaining in the stomach. That which
had been evacuated into the intestine was perhaps less hydro-
lyzed. Although the test meals given to children contained con-
siderably less peptone than those given to adults the pH values
in the recovered material were much higher than those found in
adults. No. 6 excepted. Nevertheless there is no appreciable
difference between the degree of hydrolysis in children and adults.
In each group the degree of hydrolysis of the low-molecular tri-
glycerides is large, whereas in the higher triglycerides there is a
tendency to a higher degree of hydrolysis in the case of children.
The absolute percentage of acid liberated is, however, also in
children so slight that it is of no practical importance.

Whereas the order of magnitude of the degrees of hydrolysis

* The authors are indebted to Professor Best Asueeses, M. D., for permission
to mate these investigations in his department.

Tnlilo II.

The Ilijdrolysi.i of Pure Triglycerides in the Ventricle of Adults and Infants.

THE GASTRIC LIPASE IN MAN

357

•wvs^^'^ . 1 dx^^ctc , g^atvce

^,xtv. oi

* , , „„e «"» *ir »\.’°"'‘\r*^0'”“* t^Is» '-“'f .te
v.*4 fZl i" «»* Kt ’‘i *”" '" ,

T-ot .sV«tuio'»" Tl'to W t*=®“

;S n«" -“’''“'l^ " r

Cs^T-'^'tt ’"' te V»« '^'t i'''' "'■
"0“ *«c Wv"f *= e"-'" ’"

cox^ccx^^f° cn^ctxx''® Fat

^nVivds c” c vX\csc - ,• ]^ Foi

at\d xtv _ _ ^-nvffttds

\u\

sto®"* Lt®"® Joi » ■?“"

THE GASTRIC LIPASE IN MAN.

359

material was there any sign of regurgitation from the intestine.
The only fatty substanees which seem to be digested in the stom-
ach to a considerable degree were the lower triglycerides. [From
our results with test meals the conclusion may be drawn that
the gastric lipase hardly asserts itself to a greater extent in infants
than in adults (cf. the experiment with tricaproin). It had been
stated (Pesthy 1906), that in patients suffering from achylia
the gastric lipase is not missing, which appears, too, in our experi-
ment with adult No. 6.

Our experiments on the digestion of milk fat in the ventricle
of 2 infants show that there is only a slight difference between
the splitting of the two kinds of fat during the experimental
period (25 min.). However, the degree of hydrolysis is somewhat
smaller in the case of woman’s milk fat than for the cow’s milk
fat. This was to be expected as it is well known that woman's
milk fat contains fewer of the low-molecular triglycerides than
cow’s milk, (see among others Hilditch and Meara, 1944).

We think it is not without interest that the splitting of cow’s
milk in the stomach of infants by means of gastric lipase is found
to be a little greater than in the case of woman’s milk.

Summary.

A reexamination has been carried out of the problems concerned
With the gastric lipase in man, mainly with pure triglycerides as
substrates. Both in vitro and in vivo experiments were under-
taken.

1. The optimum pH for gastric lipase towards tripropionin,
tributyrin and tricaproin in systems without the addition of
CaCla is found to be between 5.5 and 5.8. With an increasing
number of C-atoms in the fatty acids there is an evident displace-
ment of the optimum pH to the alkaline side, the optimum pH
for tricaprin, trilaurin and tristearin being 7.2, 7.3 and 7.9
respectively.

2. Addition of 0aCl2 to the enzyme-substrate systems causes
definite activation of the hydrolysis of trilaurin and tristearin at
pH lower than 7, and the optimum pH is shifted 1.5 to 2 pH
units to the acid side.

3. The relative activities of gastric lipase towards different
triglycerides have been computed. Tributyrin is split with the

360

FRm SCHGNIIEYDER AND KIHSTEK VOLQVARTZ.

greatest initial velocity, vlicteas the relative activities towards
the solid triglycerides are very small.

4. Experiments on the stability of gastric lipase at 40® C
at different pH values show that gastric lipase is a very stable
enzyme in the acid medium.

5. The lipolysis stated in our experiments in ^'ivo is according
to our opinion due to a specific gastric lipase and not to a re-
gurgitated pancreatic lipase. Experiments in vivo with test meals
containing pure triglycerides show that only the lower triglyce-
rides are split to an appreciable extent during the experimental
period of 25 min. Under similar experimental conditions there
seems to be hardly any difference betAveen the degrees of hydrolysis
of fatty substances in the stomach by means of the gastric lipase
in adults and in children. Under similar experimental conditions
in children there is a tendency to a higher splitting in the stomach
of coAV-buttcr fat than of woman's milk fat Avhen these fats are
given in test meals.

This Avork has been aided by a grant from the "Christian X’s
Fond”

Hoforonces.

Daaudsohx, H., Biochem. Z. 1912. 4-5. 284.

Gyotoku, Iv., Ibidem 1928 a. 193. 19.

— , Ibidem 1928 b. 193, 27.

Hauroavitz, F., and W. Petrou, Hoppe-Seyl. Z. 1925. IM. 68.
Hilditcii, T., and JIeara, Biochem. J. 1944. 3S, 29.

Melli, G., and M. RAmci, Fol. din. chim. microsc. 1928. 3. 169 .

Cited from Ber. gcs. Physiol. 1929. 49. 494.

OPFENnF-LAiER, C., Die Fcrmentc und ihre Wirkungen, 1925. I. 487.
— , Ibidem 1936. Suppl. Bd. I. 60.

A’. Pestiiy, S., Arch. Verdau. Kr. 1906. 12. 292.

SchOnheyder, F., and K. Yolqa^artz, Acta physiol. Scand. 1944.
7. 376.

— , Ibidem 1945. 10. 62.

From the Physiological Department, Karolinska Institutet,
Stockholm.

The Effect of Piperidine and Allied Snhstances
on Mammalian Skeletal Muscle.

RICHARD F. OHNELL.

Eeceived 30 March 1946.

By means of a special technique, now generally known as
"close arterial injection”, Brown, Dale and Feldberg (1936)
showed that acetylcholine in minute doses produces a rapid and
vigorous contraction of mammalian striated muscle. The effect
belongs to the kind of actions of acetylcholine which Dale (1914)
has termed “nicotine-like”. It was later demonstrated that various
other choline esters produce a similar effect and so does nicotine
itself (Bacq and Brown, 1937).

It has recently been shown at this laboratory (Euler, 1944,
1945 a) that piperidine is a normal constituent of human and animal
urine, being excreted — in man — in amounts of 5 — 20 mg daily.
Since piperidine is known to exert a variety of pharmacological
actions, typical of nicotine, it seemed of interest to study its
effect on the mammalian striated muscle, the more so after it
has been shown that piperidine output in urine may be considerably
increased during muscular work (Euler, 1945 b).

Apart from piperidine, a number of other hydrated pyridine
derivatives, more or less closely related to piperidine, were studie d
as to their effect on muscle.

Technique.

Cats, weighing between 2.o and 3.5 kg, were decerebrated under
n Qa^'\ tibialis anterior muscle was prepared according to Brown
(•1938). The animal was kept in a large, heated metal box serving as

3G2

RICHARD F. OHNELL.

thermostat, moist chamber and electrical screen. The muscle movement
was transmitted by means of a string to a Brown-Sciiuster spring
myograpli. The isometric contraction was recorded mcclianically on
a smoked drum, as well ns optically via a piezo-electric crystal with an
amplifier-oscillograph system, synchronously with the action potentials.
The temperature in the chamber was ns a rule kept between 33° and
37° C.

The muscle was stimulated indirectly with condenser shocks, using
silver electrodes, placed on the sciatic nerve in the thigh.

The stimulation electrodes for the dcncrvnted muscle were placed
one at the tendon, and the other in the proximal part of the muscle.

The stimulating shocks were obtained from a neon lamp stimulator
and their strength kept but slightly above that giving a maximal single
twitch. The muscle was stimulated at regular intervals (approx. 5 sec.)
and, at the injection, one impulse was omitted ad modum Baowx.

The injections lasted 1 — 2 sec. and the injected volumes were 0.25 ml.

The action currents were lead off from the tips of two enamelled
steel electrodes, attached to each other 1 — 2 mm apart. After insertion
in the muscle they were fixed b}' means of short fle.xiblc wires so as
to avoid undue displacement during the contraction. Registration with
two differentiall}' coupled amplifiers (“Triplex”, System Elhqvist,
Sweden).

The condensor-coupled amplifiers had a relatively large time constant
(approx. 2 sec.).

Solutions.

The temperature of the solutions at the time of injection was
37° C. All stock solutions kept at -f 4° C,

Locke’s or Tyrodc’s solutions, mammalian Ringer’s solution, plasma
or whole blood were used as vehicles for the examined substances.

Piperidine, as piperidinum purissimum, neutralized wth HCI.
Approximately blood-isotonic stock solution (O.iso molar).

This solution was diluted with Locke’s or Tyrode’s solution, whole
blood, etc., to the desired strength.

The stock solution was found to cause rapid haemolysis of cat blood
corpuscles. However, when made 0.155 molar with regard to NaCl,
there was no hacmol3’sis. This latter solution was compared with the
original stock solution, both at a dilution of 1 : 5 with Locke’s solution.
They gave a similar effect on the muscle as regards the size of the
elicited contraction.

In order to exclude calcium shortage, the stock solution was made
0.04 % w'ith regard to Ca CI2 in one e.xperiment. This did not noticeably
alter its effect on the muscle when used in a dilution 1 ; 5 in Locke’s
solution.

Acclplcholine. Stock solution with 1 mg acetylcholine/ml of dist.
water, acidified with HCI to pH 4 and diluted with Locke’s, Tyrode s
solution etc., immediately before the beginning of the ei^eriments.

Nicotine, as acid 1-nicotine-d-tartrate. Stock solution with O.2 /o
nicotine in Locke’s solution.

effect of piperidine and allied substances.

363

Coniine, as hydrochloride. Stock solution with 2 % coniine in dist.

^^Svartein, as sulphate. 0.150 molar stock solution .(in dist. water).
Arecoline, as hydrohromide. 0.150 molar stock solution (in dist.

The stock-solutions were always diluted in the same kind of physio-
logical salt solution as in the “blind-tests” of same experiments.

A small and constant dose of heparin in order to prevent clotting was
regularly added to the injected solutions.

Bes lilts.

1. “injection, effect”.

At an early stage in the present investigation it was noticed
that sometimes the mere injection of the salt solution, used as
a vehicle for the substances to be tested, caused a marked and
even strong contraction.

This action of plasma, Locke’s solution or other physiological
solutions, will be termed “injection effect”.

In some preparations the “injection effect” could be demon-
strated at the very first injection. In other cases, however, this
effect occurred only later on, and often became increasingly marked
during the course of the experiment. This latter type of “injection
effect” appeared, more or less pronounced, in most of the experi-
ments.

Kg. 1 A illustrates a case where the very first injection (Locke’s
solution) released a contraction.

The subsequent injection of 2.5 /ig acetylcholine in “Locke”
gave, remarkably enough, a smaller contraction than TOth “Locke”
only. After this, when “Locke” and acetylcholine (in “Locke”)
were given alternately, the latter in gradually increasing dosages
and with successively augmented effects, the former showed a
fairly constant action.

In other preparations, however, the initial injections gave no
effect. (Op. Figs 1 B and 1 C.)

Fig. 3 B shows action potentials during an "injection effect”
(Tyrode’s solution), released by the very first injection made in
the preparation concerned. The electrical activity is shorter than
at an acetylcholine injection. Still the action is of the type of a
short tetanus as shown also by the mechanogram which sometimes
exceeded in height the maximal single twitch.

364

RICHAKD F. OHNELL.

The mechanogram may retmn to the base-line, after injection
of the salt solution only, more quickly than after acetylcholine
(Fig. 1 A).

Fig. I.

A. Cat, 2.4 kg. Tib. ant. Max. indir. stimulation. Close art. injection. At Lo,
— ^Lo,: injection each time of 0.25 ml of Locke’s solution. At Ac,: 2.5 /‘g of acetjl-
choline in 0.25 ml of "Locke”. At Ac, in the same way 6 «g of acetylcholine, and
at Ac, 10 /ig of acetylcholine. The height of the contractions duo to the "Locke
injeotiona (s. c. injection effect) approximately the same on all four occasions.
Proportional to the dosage, acetylcholine gives increasing contraction heights.

B. Cat, 2.4 kg. Tib. ant. Max. indir, stimulation. Close art. injection. A t I :
Locke’s solution. No effect. II: 600 /‘g of piperidine in "Locke”. Contraction. lu:
Only "Locke'’. No effect. IV: 2.5 «g of acetylcholine in "Locke”. Contraction. V :
"Locke”. Indicated contraction.

After piperidine, slight potentiation of the single twitches; practically none
after acetylcholine.

C. Cat, 2.7 kg. Tib. ant. Max. indir. stimulation. Close art. injection, .toalogons
to fig. I B I, n, n, IV respectively, the same as in Fig. 1 B. Note: duration of pt-
peridino contraction and potentiation of subsequent twitches.

EFFECT OF PIPERIDINE AND ALLIED SUBSTANCES. 365

The tendency towards an “injection effect” appearing later
on in the experiment is, apparently, to some extent due to the
substances used in previous injections.

Piperidine was found to increase the tendency in this respect
(cp. Pig. 6 A).

In one case, where the nerve muscle transmission had been
completely blocked by spartein, contraction could still be released
by the injection of plasma or whole blood. In one experiment the
injection of only 0.05 ml elicited a small but distinct contrac-
tion.

One preparation which did not contract from the injection of
mammalian Kinger’s solution of 37° C, remained unaffected
even by the injection of “Kinger” of 0° C and other temperatures
in between.

2. Piperidine.

The piperidine effect was studied in 25 animals. In each experi-
ment repeated injections were made under varjdng conditions.

Piperidine injections released a short tetanus, as already men-
tioned, lasting between 0.5 and 2 sec. Cp. Pig. 2. The doses equalled
between 50 and 1,000 /ig. 50 /4g sometimes gave a small, sometimes
no contraction in the normal muscle.

Fig. 2. Cat. Tib. ant. Close art. injection of 230 wg piperidine. Asynchronous tetanus

for some 0.5 seconds.

As compared to acetylcholine, the piperidine myogram rises
more gradually. Pig. 3 A shows the myograms of the contractions
II and IV in Pig. 1 B, as registered with the piezo-electric device.
In other cases, the difference was still greater.

Even the relaxation is slower. Apart from Pig. 3 A, this is
noticeable, in Pig. 1 C (II), in spite of the slow rotation of the
drum.

Pig. 2 shows an irregular tetanus, released by 250 ^g of piperi-
dine in Binger’s solution. Pure mammalian Binger's solution
gave no contraction and no action potentials in this preparation.

366

EICHARB P. OHNELL.

Fig. 3.

A. Reproduction of mechanogram (piezo-electric registration) from ncetvl-
oholine (continuous curve) and piperidine contraction (dashed carvel, II and IV
in Pig. 1 B. Slower rise and fall, respectively, of the piperidine curve.

B. Cat, 2.0 kg. Tib. ant. Close art. injection of Tyrode’s solution. Action po-
tentials; the activity lasts for a good 0.1 second.

In Fig. 4r, tiie activity in what appears to he a single muscle
element is registered. This series of discharges started 2.3 sec.
after the beginning of the intitial asynchronous tetanus (=
contraction II in Fig. 1 C). Fig. 5 depicts graphically the relation-
ship between the interval of the single spikes and the time at
which the activity in question first manifested itself. Similar
curves for acetylcholine are given by Beoivk (1937 a).

Successive injections of 50, 125 and 250 ^g of piperidine gave
contraction heights with an approximate ratio of 1 : 2 : 3.

The effect of piperidine on the subsequent single twitches
varied. Figs. 1 B and 1 0 show a slight, transient post-piperidine

Fig. 4. Cat, 2.7 kg. Tib. ant. Close art. injection of 600 /<g of piperidine. Action
potentials registered. 2 — 3 seconds after the start of the piperidine effect,^ wo
reproduced series of spikes set in. The corresponding mechanogram is seen at "H
in Fig. 1 C. 1 mV'’ = 15 mm.

EFFECT OF PIPERIDINE AND ALLIED SUBSTANCES.

367

Fig. 6. Graphic analysis of Fig. 4. Abscissa: time from beginning of the serial
discharge concerned. Ordinate; intervals between the single spikes (in milliseconds).
Each point of the curve represents the arithmethio medium of five intervals.

potentiation, tvliereas Rgs. 6 A and 6 B disclose an inhibition
whicli, in the latter case, reaches maximum after about 20 sec.,
and has entirely disappeared after 2 or 3 minutes.

The potentiation is probably related to the post-tetanic poten-
tiation studied by Brown and Ettler (1938), since the piperidine
contraction is, in fact, a short tetanus. In some experiments, a
distinct potentiation was observed also after acetylcholine in-
jection.

. The inhibition, on the other hand, is most likely a manifestation
of the curare-like action exerted by nicotine and nicotine-like
substances, including also acetylcholine (cp. Eigs. 6 C and 7 D).
The curare-like action of piperidine was observed by Cushny
(1896) and others.

After denervation the muscle becomes very sensitive to piperi-
dine. Fig. 7 A shows the effect of 50 [ig piperidine in a cat weighing
2.8 kg in which the sciatic nerve had been sectioned one month
earlier. The same experiment demonstrates (at “I” in Fig. 6 C)
the effect of 0.25 fig acetylcholine, and at II, the effect of
600 fig of piperidine. The acetylcholine gave the characteristic

368

RICHAED F. fiHNELIi.

Fig. 6

A. Cat, 2.3 kg. Tib. ant. Slax. indir. stimulation about sec. apart. Close art.
injection I: injection of mammalian Ringer’s solution. No effect. II: 100 ftg of
piperidine. Contraction and subsequent inhibition of the single contractions. HI:
Again “Ringer”. This time contraction, i.e. “injection effect”. The contraction level
slightly lower than that of piperidine. No action on the subsequent single twitches.

B. Cat, 2.0 kg. Tib. ant. Slax. indir. stimulation about 6 sec. apart. Close art.
injection. I: injection of 6 /<g acetylcholine. Contraction nearly as high as the single
twitch. II: 250 /<g of piperidine. The contraction now approximately half as big
as at A. Reversible inhibition of subsequent twitches. Inhibition was maximal
approximately 20 sec. after the injection and completely disappeared after 2 or
2 minutes. Possibly the electrically induced twitches are very short tetani in
this experiment (op. Figs. 7 B and 7 D).

C. Cat, 2.8 kg. Tib. ant. Denenmted one month provio\isly. Direct stimulation
of muscle. Close art. injection. I: 0.25 ,«g of acetylcholine. "Biphasic” contraction
curve and subsequent inhibition of the single twitches. At 11, injection of 600 fig
of piperidine. Contracture, which did not relax within 3 minutes.

“disphasic” cuive (Bro'W’N, 1937), and a partial inhibition of
subsequent single twitches. Piperidine caused a contracture which
did not relax during the following 3 minutes.

3. Nicotine, coniine, spartein and arecoline.

Nicotine. (5 cats.) A dose of 15 /ig of nicotine elicited a distinct
contraction and slight inhibition of subsequent single twitches.
Pig. 7 C shows the effect of 150 /ig of nicotine: the contraction
level is higher than that of the single twitches. After this dose
indirect stimulation gave no contraction. Ringer’s solution just
before gave no “injection effect”.

Coniine. (5 cats.) Pig. 7 D illustrates an experiment where a
previous injection of Tyrode’s solution gave no “injection effect”.
250 fig of coniine, however, gave contraction and subsequent in-
hibition of the indirectly induced contractions. Each of these
contractions corresponds to a short tetanus.

Spartein. (4 cats.) This compound elicited but litte or no con-

EFFECT OF PIPERIDINE AND ALLIED SUBSTANCES.

369

Fig. 7.

A. Cat, 2.8 kg. Tib. ant. denervated one month previously. Direct stimulation
of muscle.

At "I" 50 /ig of piperidine were given: distinct contraction.

B. Cat, 2.2 kg. Tib. ant. Supramax. indir. stimulation (giving short tetani)
about 6 sec. apart. Close art. injection. I: Tyrode’s solution. Small contraction
("injection effect”). II; 140 /'g of spartein. The contraction slightly bigger than
after "Tyrode” only. Reversible inhibition of subsequent contractions. Ill: 10 (tg
of acetylcholine. Contraction bigger than at I and II.

C. Cat, 2.5 kg. Tib. ant. Max. indir. stimulation. Close art. injection. At "t”
160 fig of nicotine were injected. Contraction. Subsequent nerve stimulation no
longer caused any effect on the muscle.

D. Cat, 2.7 kg. Tib. ant. Supramax. indir. stimul.. giving short tetani at equal
intervals. Close art. injection. 1: 160 «g of coniine. Contraction and inhibition of
subsequent contractions. II; 10 fig of acetylcholine. Ill: 250 fig of acetylcholine.
After in, stimulations of the nerve give practically no contractions.

traction effect. In one test, where 2.5 of acetylcholine gave
a distinct contraction, 0. 9 mg of spartein gave neither contraction
nor any effect on subsequent single twitches. In another prepara-
tion, 0.25 mg spartein gave some contraction (hard to interpret,
however, owing to “injection effect”), and then a total neuro-
muscular blockage. In other experiments, again, a reversible
inhibition of subsequent contractions was observed (Fig. 7 B).

Arecoline. (3 cats.) Doses of 0.6, 1.2 and 6.0 mg respectively,
produced no contraction in a preparation with normal sensitivity
to acetylcholine and piperidine. Subsequent single twitches were
inhibited to a varying degree, the inhibition being roughly pro-
portional to the dosage. Also in other experiments arecoline was
inactive with regard to the contraction effect.

24 i60215. Acta phys. Scandinav. Vol. 11.

370

lUCUARD •]?. Oil NELL.

Discussion.

The fact that the mere injection of a physiological solution
(Locke’s solution, whole blood, etc.,) may elicit a muscle contiac-
tion is of some interest, primarily perhaps, as a source of error.
It should be emphasized that large amounts of fluid are not
required for the producing of this effect which could be observed
after as small a volume as 0.05 ml, and without using a particularly
high injection velocity. After total neuromuscular blocking (by
means of spartein), small contractions could still be released by
plasma or whole blood. This suggests that the point of attack
of the “injection effect” may be further peripheral than the motor
neuron.

The closer nature of this effect (mechanically or chemically
— variations in pH? — elicited stimulation) was not investigated.

The demonstration of the capacity of piperidine to induce a te-
tanus has lead to the ascertainment of still another body-specific
substance capable of releasing a contraction in mammalian skeletal
muscle.

Analogously to acetylcholine and nicotine (Langley, 1913/14,
Buchtal et al. 1942, and others), the point of attack of piperi-
dine in the muscle is probably the end-plate.

As regards the observations made by Buchtal ei al. (1944)
concerning a contraction released by adenosine triphosphoric acid
and related compoimds, this effect is apparently connected to
structural changes of the myosine molecule.

Compared with acetylcholine, an about 200 times larger dose
of piperidine was needed in order to cause the same contraction
height. However, such a comparison seems somewhat objectionable
since piperidine gives a slower effect regarding the period of
contraction as well as the relaxation time.

The capacity of piperidine to- produce contractions in the
isolated rectus abdominis of the frog (Euler and Domeij, 1945)
is not necessarily analogous to the effect described in the present
paper. Thus, they behave differently with regard to acetylcholine
(cp. Brown, 1937). Another apparent discrepancy between
amphibian and mammalian muscle is to be met with in the
present paper: arecoline fails to cause contraction, in spite of the
fact that it has been found capable of this effect in the muscle
of frog.

EFfECT OF PIPEniDINE AND ALLIED SDBSTAKCES.

371

The action of piperidine on various biological objects has always
been found to be weaker, weight for weight, than nicotine.
Apparently this also applies to the capacity of producing a con-
traction effect in mammalian skeletal muscle. The relation between
nicotine and piperidine with regard to the effect on the blood
pressure equals 20 : 1 (Dixon, 1924), on the rectus abdominis
muscle of the frog 200: 1 (Euler and Domeij, 1946). From the
present experiments, the ratio may be given as approximately 10: 1.

Coniine causes contraction with inhibition of the subsequent
contractions whereas spartein gave slight or no contraction in the
doses employed. However, there was at times a distinct, some-
times reversible, inhibition of the subsequent twitches vdth
spartein. This is apparently a curare-like effect, as a contraction
could still be elicited by direct stimulation (?. e. “injection effect”
still demonstrable).

I am greatly indebted to Professor U. S. v. Euler for suggesting
the problem and for his inspiring interest in the progress of the
work.

Summary.

1. Piperidine in doses of 50 /ig or more produces a short tetanus
in the tibialis anterior of the cat, when applied ad moduni Brown
(1938), i. e,, “close arterial injection”.

2. The effect of subsequent maximal single twitches is either
a transient potentiation or else inhibition (the latter sometimes
reversible).

3. The myogram of piperidine has a slower rise and fall than
the corresponding acetylcholine curve.

4. Denervated muscle is highly sensitive to piperidine. It
causes contraction and, in larger doses, contracture.

5. Injection of Locke’s solution, plasma, or some other physio-
logical solution may sometimes produce a short tetanus (“Injec-
tion effect”).

6. Nicotine and coniine give contraction and subsequent inhibi-
tion of indirectly induced contractions.

7. Spartein and arecoline give slight or no contraction, but
distinct inhibition of indirectly induced contractions.

372

RICHARD F. SHNELL.

References.

Bacq, Z. M., and G. L. Brown, J. Physiol. 1937. 89 . 45.

Brown, G. L., Ibidem, 1937. 89 . 220 and 438.

Brown, G. L., Ibidem. 1938. 92 . 22 P.

Brown, G. L., H. H. Dale and W. Peldberg, Ibidem 1936. 87 . 394.
Brown, G. L., and U. S. v. Euler, Ibidem 1938. 93 . 39.

Bdchthal, E., and G. Kahlson, Acta Physiol. Scand. 1944. 8 . 317.
Buchthal, F. and J. Lindhard, Ibidem 1942. 4 . 136.

CusHNY, A. R., J. Exp. Med. 1896, 1 . 202.

Dixon, W. E., Heffter’s Handb. exp. Pharmakol., 1924. 2 : 2 .

Euler, IT. S. v., Acta Physiol. Scand. 1944. 8 . 380.

Euler, U. S. v.. Nature 1944. 154 . 17.

Euler, U. S. v., Acta Pharmacol. 1945 a. 1 . 29.

Euler, U. S. v., Acta Physiol. Scand., 1945 b. 9 . 382.

Euler, U. S. v., and B. Domeij, Acta Pharmacol. 1945. 1 . 263.
Langley, J. N., J. Physiol. 1913/14. 47 . 159.

From the Department of Physiology, University of Lund.

Effect of Minute Amounts of Bjirium
on Cardiac Muscle.

ADAM DEDTSCH' and GDNNAR LDNDIN.
Received 8 April 1946.

In the course of investigations on striated skeletal muscle
(Buchthal et al. 1944), tve have examined the effect of adenosine
triphosphate on small bundles of cardiac muscle fibres without
automatic activity. Former investigations have established that
adenosine triphosphate in moderate concentrations has no effect
on normal cardiac muscle, while contraction tension is claimed to
increase in “hypodynamic” preparations (Lindner and Rigler
1931). Large amounts are found to release block and other dis-
turbances in the conducting system (Gillespie 1934).

In our first experiments adenosine triphosphate (ATP) applied
to small bundles of cardiac muscle had a striking effect, releasing
automatic activity of several minutes duration. This effect, how-
ever, could not be reproduced regularly and it was soon realized
that it was due to an accidental contamination by small amounts
of Ba salts of the ATP solution used in the first experiments.
This made us to investigate the effect of Ba salts in different con-
centrations.

Apart from ATP other phosphorus compounds have been
examined with regard to their action on cardiac muscle.

Method.

The experiments were performed on small approximately parallel
threaded muscle bundles O.i — 0..3 mm thick and 1.5 — 2 mm long, from
the cardiac ventricle (Rana esculenta), prepared in an ice-cooled Ringer
^iution of pH 7.3. The Ringer solution contained 0.67 g NaCl, O .02 g
RCl, O.oi g CaClj. 6 H 2 O and O .02 g glucose in 100 ml. Oxygen content

' Working with a grant from the Stvedish Medical Research Oov,ncil.

374

ADAM DEUTSCH AND GUNNAR LUNDIN.

and constant pH were ensured by passing a stream of 99 per cent
0, and 1 per cent CO^ through the solution and by adding a suitable
amount of NaHCOj. Normal colloid osmotic pressure was maintained
by the addition of 3 per cent Dextran (Gronwadl & Ingelman,
1945) to the solution. Determination of the mechanical tension has been
performed with a condenser myograph according to Buchthae (1942).
This consists of a fixed and a movable condenser plate. The movable
plate is connected with a pair of micro-tweezers which hold one end
of the muscle bundle. IVhen the muscle contracts the distance between
the two condenser plates is diminished and changes in capacity are
induced which are registered by means of a high frequency circuit,
d. c. amplifier and electrostatic oscillograph. The changes in electrical
tension due to variations in capacity are proportional to the A'ariations
in mechanical tension of the cardiac muscle.

The muscle was placed in a chamber containing O.is ml Ringer
solution and was held by two pairs of micro-tweezers which were used
as stimulating electrodes. To reduce polarisation tweezers made of
silver were used. Excitability and strength-duration curves were deter-
mined with rectangular current pulses of variable duration from a sti-
mulator of multivibrator type. Control of the electrical a. c. conductivity
between the electrodes ensured that changes in excitability were not
due to changes in resistance in the solution -f muscle. The measuring
frequency was 1,000 cycles per sec. For details we refer to a previous
paper (Lundin 1944).

The experiments were performed at 20° C, the temperature being
checked thermoelectrically.

Preparation of Substances.

Adenosine triphosphate (ATP). The ATP was prepared from rabbit
muscle as the Ba salt by the method of D. M. Needham (1942), and as
the neutral or acid Ba salt and the Ca salt by the method of S. E.
Kerr (1941). No Ba salts and only Ba-free reagents were used in the
preparation of Ca ATP.

A solution of Na ATP was prepared from the Ba salt by addition of
the calculated amount of sodium sulphate to the Ba salt either sus-
pended in water or dissolved in dilute hydrochloric acid.

The Ca ATP was converted into the Na salt by the addition of the
calculated amount of sodium oxalate to the solution of Ca ATP in
dilute hydrochloric acid. Both the isolated Ba and Ca salts and the
final Na ATP solutions were analysed for N and for total and 7’ P
according to the method of Fiske and Subbarow (1925) in the modi-
fication of ScHEEL (1936). The purity of each substance was at least
98 — 99 per cent.

EFFBCr OF MINUTE AMOUNTS OF BARIUM ON CARDIAC MUSCLE. 875

• H»0 requires: P, 8.5 %, N, 11.5 %. A solution of sodium creatine plios-
phase was prepared by precipitation of Ba with the calculated amount
of sodium sulphate from the solution of the Ba salt in water.

Ace(i/l fhosfliate. The substance was prepared according to Lynen
(1940) as the silver salt (P, 8.9 %; calc, for GjHjOjPAg : P, 8.8 %).
A solution of the sodium salt was prepared by addition of the calculated
amount of NaCl to the suspension of the silver salt in water.

Sodium triphosphate (NasPaOjo'bHoO). The substance used con-
tained c. 85 per cent NajPjOm • SHjO and 12 per cent sodium ortho-
phosphate.

Sodium pyrophosphate, sodium orthophosphate, sodiwn metaphosphate
and the different barium, calcium, magnesium and strontium salts used
were of analytical purity.

All substances were applied in iso-osmotic solutions, by replacing part
of the NaCl -j- water in the Ringer solution by an equivalent amount
of staple solutions, of the substances tested adjusted to pH 7.8.

Results.

1. Effect of Ba. Of different inorganic ions investigated Ba
showed a quite unique effect independent of its application as
chloride, nitrate or acetate. BaClj regularly initiates automatic
activity, the threshold concentration being 0.3*10"’^ mol/ml.
The absolute amount of Ba added is 0.7 /<g. The automatic ac-
tivity is preceded by definite changes in irritability, which also
manifest themselves in the strength-duration curve (Big. 1).

Ordinate: strength of stimuli in arbitrary units.
Abscissa: time in ms.

376

ADAM DEUTSCH AED GUKNAR LtTNDIK.

Fig. 2. Strength-duration curve of cardiac muscle in Ringer solution { — 9 — 9 — )
and after application of Na triphosphate ( j i — ).

Ordinate: strength of stimuli in arbitrary units.

Abscissa; time in ms.

Witlx a duration of the stimulus of 2 ms the threshold decreases
approximately 30 per cent in the course of 2 — 3 nainutes (Fig. 3)
follovred by automatic actmty.

2. Stronthm also initiates automatic activity though the
threshold amounts necessary (5.10~® mol/ml), are approximately
fifty times higher than vdth Ba. Calcium and magnesium chloride
in the same concentrations have no effect on excitability, but
the addition of CaCh (10“® mol/ml) increases the strength of
contraction up to 45 per cent.

3. Adenosine triphosphate prepared in different ways and
applied as the sodium salt in concentrations of 1.8 — 3.6.10"*’
mol/ml had no effect on the excitability and contractility of car-
diac muscle, while highly active in experiments on skeletal muscle.

4. Application of creatine phosphate in concentrations of
3.6 — 7.2.10"® mol/ml causes a decrease in the mechanical
response of approximately 50 per cent, five minutes after addi-
tion of the solution. The simultaneous increase in excitability
(duration of stimuli 2 ms) amoimts to approximately 30 per cent.

5. Acetyl phosphate in concentrations of 3.6 — 7.2.10"® mol/ml
is without effect on excitability and contractility.

6. Sodium triphosphate and pyrophosphate in concentrations
of 3.6 — 7.2.10"® mol/ml cause in the course of a few minutes a
considerable reduction in contraction tension, and in most cases

EFFrCT or MIKUTK AMOUNTfS OK liAlUUM ON CAIUtlAC MUSCU:. 377

contrnctiHty disappears completely. This effect is reversible niid
after Avnsliing out with Kinger .eolution for 1 — 2 minutes, the
original strength of contraction is quickly restored. Excitability
after triphosphate for a duration of the .«tiinulus of 2 ms increase.s
25—70 per cent depending on the concentration applied. The
increa.sc is most pronounced with short stimuli ns .seen from the
strength-duration curve in Fig. 2.

I'ip. 3. tAcitnliility of cnrtlinc nui'Ho jin'pnrntion oft^'r jn/I0,000 ( — • — ® — ).
ra/30.000 ( — I — 4 — ) nnil ni/2.'i,Ofrfi (— (3 — 0“-~)I!nri,. The nrrow ilmotcs start

of iiiitoiimtir' iictirity.

Ititrnlinn of utinitili 2 ni«.

Ordinate: Htn-netli of a'.itntili in nrln’trnry miils.

AbsciMa: tiiiu; in jninnfrs.

7. Sodium nu:t(i])!iosphatc. in concent rntion.s of 3.G — 7.2.10"®
niol/inl .also cau.sc.s a .strong decrease in the strength of contraction,
while orthophosphate affects mechanical tension only slightly.
Excitability for stimuli of 2 ms duration is im])roved after py-
rophosphate and mcla])hosj)ha1o and is more pronounced with
still shorter stimuli of 0.2 ms duration, when even orthojdvosphatc
improves cxcit ability.

Discussion.

Although the stimulating effect of 3?n on smooth and striated
muscle is well established, observations with regard to its action
on cardiac muscle arc rather contradictory. It has often been
compared with the action of digitalis as many authors describe
its effect ns a decrease in frequency and an increase in mechanical
contraction tension. There is unanimity with regard to the effect
of large amounts of Bn, ovoquing a systolic contracture. Small
doses arc stated to increase frequency (Salant & Klkitman,

376

ADAM DEUTSCH AKD GUNNAK LEKDIN.

1922) or to have the reverse effect (Liotta, 1924:, Tocco-Tocco,
1924). It has been used therapeutically in cases of auriculo-
ventricular dissociation (Cohn and Levine, 1925), where it ap-
parently has a regulating effect.

Kisch (1927) has investigated the action of the alcaline earth
cations on the initiation of impulses in the frog’s heart and found Ba

more active than the other alcaline earths
in facilitating the physiological stimuli.
Our investigations establish the specific
action of the barium ion on cardiac muscle,
releasing an automatic activity of several
minutes duration, when applied in minute
amounts.

Our results with adenosine triphosphate
indicate that ATP in concentrations which
are highly active in skeletal muscle is with-
out any effect on the contractility and
excitability of the cardiac muscle.

Since in some of our earliest experiments
an apparent heart action of ATP was due
to a contamination with minute amounts
of Ba, we have examined whether the
antagonistic effect of ATP to acetylcholine
(Abdon, 1942) also can be produced by small amounts of BaCb.
We find in the Straub heart preparation that BaCU in a concen-
tration of IQ-' mol/ml counteracts the effect of acetylcholine
on the frogs ventricle (Fig. 4).

Summary.

1. Barium salts in minute amounts initiate automatic activity
in the cardiac muscle, the threshold concentration being 0.3 *10“'
mol/ml. The automatic activity is preceded by a definite improve-
ment in excitability. This effect is specific to Ba; Ca and Mg are
inactive and for Sr, the threshold concentration is approximately
fifty times higher than for Ba.

2. Adenosine triphosphate has no effect on cardiac muscle in
concentrations of 1.8 — 3.6*10”® mol/ml.

3. Creatine phosphate in concentrations of 3.6 — 7. 2* 10"®
mol/ml causes a decrease in the mechanical response of c. 50 per
cent.

• 4 ■ V ';.' i'-l

lit'

Fig. 4. Antagonistic effect
of barium to acetylcholine
on the Straub heart pre-
paration. Arrow 1 denotes
the addition of acetylchol-
ine 1 : 10% then washing out
with Ringer solution and at
arrow 2 addition of BaCl.
(10“' mol/ml) and arrow 3
renewed addition of acetyl-
choline 1 : 10'.

EFFECT OF MINUTE AMOUNTS OP BAEIUM ON CAEDIaC MUSCLE. 379

4. Inorganic triphosphate, pyrophosphate and metaphosphate
in concentrations of 3.6 — 7.2-10““ mol/ml cause a considerable
reversible fall in contraction tension of cardiac muscle. Ortho-
phosphate produces a slight decrease in the strength of con-
traction.

5. Barium chloride in a concentration of 10“' mol/ml counter-
acts the effect of acetylcholine on the frogs ventricle in the Straub
heart preparation.

The authors Avish to express their gratitude to hlr. F. Buchthal,
M. D., for valuable advice.

References.

Abdon, N. 0., Om Kreatinfosforsyrans och Adenosintrifosforsyrans
Betydelse for de parasympaticomimetiska Faimakas Verkan.
Thesis. Lund. 1942.

Buchthal, F., Det Kongl. Danske Vidensk. Selskab Biol. Medd.
1942. 17. 2.

Buchthal, F., A. Deutsch, and G. G. Knappeis, Acta Physiol.
Scand. 1944. 8. 271.

Buchth.al, F., and B. Folkow, Acta Physiol, Scand. 1944, 8. 312.
Oohn, a. E., and S. A. Levine, Arch. Int. Med. 1925. 86. 1.

Deutsch, A., M. G. Eggleton and P. Eggleton, Biochem. J. 1938.
82. 703.

Fiske, H. C., and Y. Subbarow, J. biol. Ohem. 1925. 66. 375.
G1LLE.SPIE, I. H., J. Physiol. 1934. 80. 345.

Gronwall, a. and B. Ingelman, Nature, Lond. 1945. 155. 45.

Kerr, S. E., j. biol. Chem. 1941. 189. 131.

Kisch, B., Naunyn-Schmiedebergs Arch. Path. 1927. 124. 210.
Lindner, F. and R. Rigler, Pfliig. Arch. ges. Physiol. 1931. 226. 697.
Liotta, D., Arch, di farmacol. sper. e sci. aff. 1924. 87. 111.

Lundin, G., Acta Physiol. Scand. Supplement XX. 1944.

Lynen, F., Ber. d. dtsch. chem. Ges. 1940. 78. 367.

Salant, W. and N. Kleitman, J. pharmacol. 1922. 20. 247.

ScHEEL, K. C., Z. analyt. Chem. 1936. 105. 256.

Tocco-Tocco, L., Arch. int. pharmacodvn. 1924. 29. 489.

Zeile, K., and G. Fawaz, Hoppe-Seyl. Z. 1938. 256. 193.

From the Pediatric Clinic of Korolinska Institutet at Norrtull Hospital,

Stockholm.

The Principle of Evacuation of the Stomach in
Infiints and Premiitnres.

A non-roentgen ologic study.

STEPHAN VENDEE.

Beceived 16 April 1946.

For several decades the motor function of the stomach has
been the object of comprehensive investigations (cfr. Catel
1936), but despite the minute observations of many experienced
investigators the separate results will be seen to differ so much
from each other that it is still possible only to speak about the
emptying of the stomach in the most general terms. The time of
evacuation varies very considerably in different individuals, and
even in the same individual on repeated examination. A fairly
great variance is found between different groups, e. g. when
comparing a number of works of the medical literature, as did
Bousloug (1935).

Therefore, Smith (1945), is fully justified when he writes as
follows: “No uniformity is to be expected. The stomach empties
with unpredictable variability”, but it is difficult to imagine
that this should really apply to infants who in most of their
other vital functions display so marked a regularity. It has been
tried to explain the variations found from the affectibility of the
stomach by its reflex and hormonal regulation. On the other hand,
in Catel’s statement (vol. I, page 121) the following description
of the gastric peristalsis is found: . . . “eine Prazision der rhyt-
mischen Tatigkeit, die Uberrascht . . . kaum um Bruchteile diffe-
rierend” !

EVACDATIOK OF THE STOMACH.

381

Nearly all investigations are roentgenologic. In most cases a
contrast medium vas employed, but a few investigators have
wanted to safeguard against possible methodic errors caused by
tills examination and made the transillumination without using
barytes, as did Behrexdt {1923) and Bessau, Rosenbaum and
Leichtentritt (1921). The predominant use of the roentgenologic
method must be attributed to the fact that it has been believed
that all factors changing the natural course of the evacuation could
be avoided: The process can be followed without interference and
the peristalsis observed, and it can be seen directly when the last
residue leaves the stomach; and, if desired, these impressions
can be fixed on a roentgen film. With some practice a certain
perception of the course is obtained from the extent and density
of the roentgen shadow: In some cases it will be seen that to
begin with the stomach empties more rapidly, later on far more
slowly while, in other cases, the reverse seems to be the case.
The flaw of the roentgenologic method is that, besides making
the peristalsis visible (and showing, when the first portion of
the meal enters the duodenum, if a barium meal has been given)
it gives only one exact item of information; The time Avhen the
stomach gets empty — and that with so great variations that,
in spite of everything, the method must be suspected of being
encumbered with great sources of error. It gives no useful in-
formation about the entire intervening part of the evacuation
curve, and cannot do so either until methods have been prepared
like that used in determination of the volume of the heart
in vivo.

Both evacuation curves have their adherents among physio-
logists. The divergence of opinions is perhaps best illustrated by
the following quotations: Stabling; "As emptying proceeds,
the rate of evacuation is readily slowed down by influences,
probably reflex, but also in part hormonal . . and by Wigger
(1934): “The emptying rate of the stomach increases progressively
from the onset to the completion of digestion.”

Only through quantitative determinations in the course of the
evacuation will it be possible to elucidate which type of emp-
tying is the right one. Such determinations are found only ex-
ceptionally in the literature, but Wilson’s statements afford
an example. Graphic representation of his table values conveys
the impression that the nature of the food itself determines the
type of evacuation; Raw egg-white represents the former — the

382

S. VENDEI..

“concave” curve; -wliereas egg-yolk with bacon results in tke
“convex” shape. By fluoroscopy an estimate was made of the
percentage of the barium remaining in the stomach after l‘/j, 3
and 4^/2 hours.

As the times of evacuation in infants displayed so wide limits
of variation, even a rough method with comparatively great ex-
perimental errors Avill be justified if only it elucidates one new fea-
ture of the evacuation of the stomach, be it either so that the
variations have been due to the natural conditions or have simply
reflected hitherto unnoticed methodic errors in the roentgen
examination.

In the present investigation our aim has been to follow the
entire course of evacuation, and not to be content with deter-
mining the initial and terminal points of the curve. The method we
chose was simple use of an ordinary stomach tube (size 12 to 14)
and a graduated Record syringe of 50 or 100 cc. The rate of evacua-
tion is followed by emptying the stomach at regular intervals
(by means of suction with the syringe), measuring the quantity
removed and then injecting it again into the stomach until it is
time again to remove it through the stomach tube. Unfortunately
a permanent tube cannot be used in infants, the tube has to be
introduced again for each determination. It would moreover be
desirable to be able to empty the stomach completely and refill it
without changing its volume in the course of the procedure, to
avoid stimulation of the peristalsis. This is possible only when
using a Miller-Abbotfs double channel tube, which is of too
coarse a caliber for this examination in small infants. Besides it
proved possible to arrive at satisfactory results without these pre-
cautions and then to reduce the number of necessary introduc-
tions of the tube, so that disturbances originating from the use
of the stomach tube were completely eliminated from the result
of the experiment.

The basis of this communication is a series of 110 experiments,
carried out on 27 children, aged 1 week — 11 months. Of these 3
were examined shortly after gastrointestinal disturbances, and in
5 cases pharyngitis was present. In the rest of the children only
disease of minor importance to gastric function, or no disease at all
was present. A number of formulae were given; mother’s milk,
acidified milk, etc., to exemplify any commonly used formula,
but the material hitherto collected being by this reason too small
and too heterogeneous to give exact informations of the various

EVACCJATIOX OF THE STOMACH. 383

emptying times, yet will allow certain preliminary conclusions
about tbe process of evacuation:

1) In most infants tbe stomach empties completely regularly,
despite tbe disturbing influence of the stomach tube; the eva-
cuation proceeds with the precision of clockwork till at least
s/s, or sometimes “/jo (or the whole) of the meal, has left the

Fig. 1.

stomach, then its emptying rate decreases somewhat. With this
qualification it can be said that the stomach empties a constant
amount per time unit almost during the whole of the meal.
In graphic representation it means a rectilinear curve. If this
line is elongated (extrapolated) till it intersects the zero line,
the result will be a time which might be termed the ideal emptying
time for the meal in question and wliich it would be possible to
express in minutes. The number of minutes required by the stom-
ach to empty 100 cc is an exponent of the emptying rate, but in
comparisons between the emptying times of children at different
ages the inverse proportion, i. e. the amount emptied per hour, is
advantageously used as emptying coefficient (= cj. The quantity
Ct varies inconsiderably as long as it is not determined for very

384

S. VENDEL.

large or very small meals: the linear curve applies vrithin -wide
“normal limits”. This conception of the emptying process is seen
to be in full harmony vdth the results of Crider ■ and Thomas’
investigations: "After larger meals the emptying time is greater
in most subjects than after small meals; e. g. doubling a small
test meal may increase the time by 17 per cent, and trebling it

may increase it by 38
percent.” An approximate
prolongation of 100 and
200 per cent should have
been less confusing than
these 17 and 38 per cent,
but we have to reckon in
most cases with a comae
of emptying, different
from the “ideal”. The
reasoning is elucidated
best by the curve (fig.

2). In this connection also
Bouslougs statement —

4 • ' • • ■ 1 • •

a cc . 1 -fc minutes

2»Q:. t+17pCt.

..zn

3 "a; 1 t + 38f>Ct.

Fig. 2. Result of Cridek and Thomas’ in-
vestigation grafically reproduced; a curve resem-
bling the last part of our curves.

“it would appear that feeding before the stomach was empty
did tend lo lengthen the emptying time” finds its natural ex-
planation. The same holds good of the observation that debile
infants would empty their stomachs more rapidly than normal
infants — they get smaller meals!

2) Ct varies fairly moderately in the individual child from time
to time, there is a somewhat greater variation between individuals
within the same age group. Our material does not yet allow the
exact determination of the variance (o).

3) On comparison between the mean figure for Ct of the dif-
ferent age groups it is seen that this constant increases reg-
ularly month after month for each kind of food in proportion
to the growth of the child. If c^, is, therefore, divided by the
normal weight for the age in question, a quotient results which
is nearly constant all through infancy for a definite sort of food
or milk mixture, whereas in prematures it is found to be de-
creased in proportion to the debility of the child, expressing
its functional insufficiency.

4) This reduced emptying coefficient is of typical magnitude
for each kind of food. The ratio between these coefficients is
found to be fairly unaltered on determination of the c^ of the

EVACUATION OF THE STOMACH.

385

single milk mixture in the individual child. Bearing in mind the
experience of previous researchers about emptying times, it is
astonishing to see how moderate the deviations in this ratio arc
as compared with the mean figure.

Exact, statistically tenable determinations of these figures are
being prepared and will be imblished in a subsequent paper,
in which reasons of individual variations will also be discussed;
the paper •will also comprise investigations into the conditions
which, in our opinion, determine the different emptying coefficients
for the various kinds of food in infants.

Summary.

By roentgen only the initial and final points of the evacuation
curve of the stomach can be determined exactly enough. To follow
the entire course of evacuation another technic was applied: the
volume of gastric contents being measured at regular intervals
by aspiration through a stomach tube. It was demonstrated that
the stomach in infants empties with extreme regularity, follow-
ing mostly a rectilinear curve, the empt}dng rate depending up-
on age and the constituents of the meal, and the variations tend-
ing to be much more confined than previously accepted.

Hoforoncos.

Alwens W. & J. Husler, Fortschr. Btgstr. 1912. 19. 183.
Behrendt, H,, Jb. Kindcrhk. 1923. 102. 291.

Bessau, G., S. Rosenbaum and B. Leichtentritt, Jb. Kinderhk.
1921. 95. 123.

Best, C. H. and N. B. Taylor, The physiological basis of medical prac-
tice. 3rd cd., Baltimore. 1943.

Bousloug, J. S., J. Fed. 1935. G. 234.

Cannon, W.B., The mechanical factors of digestion. New York 1911.
— , Digestion and health. New York 193G.

Catel, W., Die normalo und pathologische Physiologic der Bewegungs-
vorgfinge im gesamten Vcrdauungskanal. Leipzig 1936.

Crider, J. 0. and J. E. Thomas, cit. from Evans 1945 and Wiggers 1944
Evans, C. L., Principles of human physiology. London 1945.

Taylor, R., Amcr. Dis. Childr. 1917. U. 233.

Theile, P., Z. Kinderhk. 1917: 15. 152.

Wiggers, C. J., Physiology in health and disease. Philadelphia 1934
and 1944 (4th ed.).

Wilson, M. J., Can. J. racd. Ass. 1931. 25. G85.

— , Arch. Int. Med. 1929. 44. 787.

25 — i60215. Ada pliys, Scandinav. Vol.ll.

ACTA PHYSIOLOGICA SCANDINAVICA

VOL. 11. SUPPLEMENTUM XXXIII.

From the Histological Department, Karolinska Institutet
(Head: Professor GOsta HSgggvist), and the ^fedical Clinic, Karolinska sjukhuset
(Head: Professor Nanna Svartz), Stockholm, Sweden.

CONTRIBUTIONS TO
THE KNOWLEDGE OF THE EFFECT
OF EXOGENOUS INSULIN ON
THE GLYCOGEN STORAGE
OF NORMAL ANIMALS

AND A SURVEY OF FACTORS NORMALLY
INFLUENCING THAT STORAGE

AKE SWENSSON

Stockholm 1945

Till mina fordldrar

Contents

Preface 7

Introduction 9

Part I. Methods.

Chapter 1 . The glycogen analysis method 15

A. Description of the method 15

B. Testing of the method 17

C. Other redueing substances than glucose possibly found in the

hydrolysate 24

Chapter 2 . Statistieal methods 26

Chapter 3. Koutinc of the experiments 28

Part II. Survey of factors normally influencing
the glycogen storage.

Chapter 4. The glycogen reserves of the body 33

Chapter 5. Distribution of glycogen in the glycogen reserves 36

A. The glycogen distribution in the liver 36

B. The distribution of glycogen in the skeletal museles 39

Summary 42

Chapter 6 . Mode of killing and postmortal glycogenolysis 43

I. Mode of killing 43

H. Postmortal glycogenolysis 46

A. In the liver 47

B. In the muscles 51

C. The initial postmortal glycogenolysis 53

Summary 57

Chapter 7. The bearing of the animals’ age and sex on the amount of

the glyeogen depots 59

A. Sex 59

Summary 53

Chapter 8. The bearing of the diet and the length of the fasting

period on the glycogen depots 70

I. The bearing of the diet 70

n. The effect of fasting 74,

Summary 81

Chapter 9. Cyclic changes in glycogen content and variations due to
temperature 82

A. Diurnal variations - . . - 82

B. Seasonal variations 83

C. The effect of temperature 84

Summary 85

Chapter 10. Further factors which affect the glycogen content in
normal animals 87

Part IIL The effect of exogenous insulin on the glycogen
storage of normal animch.

Chapter 11. Survey of reports in the literatur on the effect of insulin
on the glycogen storage of normal animals 93

A. Different ways of studying the effect of insulin on the glycogen

storage 93

1. In vitro experiments with tissues and tis.sue slices 93

2. Perfusion experiments on surviving tissues 94

3. Attempts to draw conlusions regarding the effect on the liver

glycogen from the variation of the blood sugar after insulin
injection 97

4. Studies on the effect of insulin on the glycogen stores in

intact animals 98

B. Experiments with series of insulin-treated animals and control

animals 106

Summary 109

Chapter 12. Effect of insulin on the glycogen stores in mice 110

Chapter 13. Experiments on rats 140

Chapter 14. Experiments on rabbits ' 144

General survey of results 148

Bibliography 152

PREFACE

The present investigation was carried out at the Histological
Department of Karolinska Institutet at Stockholm during the
years 194'2 — 1945.

My thanks are due to Professor Gosta Haggqvist for his
interest throughout the investigation and for permitting me
to do my experiments at liis Department.

I am greatly indepted to Professor Nanna Svaetz for her
stimulating interest in my work and for valuable criticism.

I am very greatful to Laborator Erik Joepes and Prosektor
Hjal:mae Holmgren, who always followed my work with
the greatest interest. Their stimulating and construetive criti-
cism has been my greatest stimulus throughout the investiga-
tion.

It is a pleasure for me to thank my friend, Docent Beor
Rexed, who has always been ready for valuable discussions
and who has given me great assistance in the final adjustment
of this book.

For valuable advice respecting the statistical treatment of
the material I owe a dept of gratitude to Docent Leonard

Goldberg.

My thanks are due to Mr Grenville Grove for his
conscentious translation of this paper.

The technical work in this investigation has been carried
out with great thoroughness and untiring zeal by Miss Ann-
Makie bang and by my wife Annalisa Swensson, to
whom I wish to express my gratitude, as also to Miss Maj
Beeghman, who prepared the diagrams.

The cost of this investigation has been defrayed by grants
from Stiftelsen Therese och Johan Anderssons Minne, Medi-
cinska Prisgruppens Sarskilda Fond, Kungliga Vetenskapsaka-
demiens Lindahlfond and Apoteksvarucentralen Vitrum,

Stockholm, October 1945.

Ake Swensson

INTRODUCTION

Even a superficial study of the reports in the literature regard-
ing the effect of insulin on the glycogen content of the liver and
muscles shows that there is much confusion on this subject. Ever
since Claude Bernard in 1855 detected glycogen in the liver,
this substance has been subjected to much research and, as soon
as serviceable insulin preparations were obtained, a number of
investigators began to study their effect on the liver and muscle
glycogen. Since the beginning of the twentieth century glycogen
metabolism had in fact assumed a central place in metabolism
research. As investigators at first did not realize the amount of
the biological variation, they contented themselves with very few
experiments, which is one of the reasons why the literature on
the subject is full of contradictory statements. As the importance
of even minute variations in the experimental conditions was not
realized, incommensurable experiments were compared, and the
discussion was often based on comparisons with the results of
other investigators, even under conditions where they were not
applicable. Other publications on the subject are marred by
fundamental errors in the discussion and appraisement of the
results.

Despite the immense amount of labour bestowed on the study
of this question, however, it has still been found impossible to
arrive at any consensus of opinion.

The views expressed in the literature in regard to the effect of
insulin on the glycogen depots of the body in fact show consider-
able discrepancy. Thus, v. Meyenburg (1924)) states: j.Zusam-
menfassend mochte ich also sagen: Es scheint sicher gestellt dass

unter der Insulimvirkung eine Glykogenstapelung in der Leber
stattfindet, wenigstens bei einigen Tierarten. Ob alle bei der
Insulinhypoglykamie aus dem Blut verschwindende Zucker in der
Leber konzentriert wird, bleibt dagegen unsicher.»

Macleod (1924<) states: »There can be no doubt that insulin
causes a rapid reduction in the amount of glycogen in the liver
and may cause it almost, if not entirely, to disappear from the
muscles.!

Grevensttok and LAQtrExm (1925 b), after a thorough survey of
the literature of that time, consider it proved that insulin induces
»bei normalen, gut genahrten, wie hungernden Tieren Abnahme
des Glykogens, Zunahme des Glykogens bei pankreasdiabetischen
Tieren, aber wahrscheinlich auch bei normalen, wenn gleichzeitig
mit dem Insulin Zucker gegeben wird^. They add that they cannot
state for certain whether insulin has any effect on the glycogen
content of the muscles.

Isaac and Siegel (1928) sum up their views as follows; sUnter
Einwirkung zugefiihrten Insulins wird Zuckerverbrennung und
Glykogensynthese in der Muskulatur beschleunigt!. As regards
the liver, they state: sDas Wesen der Insulinwirkung, soweit die-
selbe heute erklarbar is, besteht in einer Beschleunigung des ge-
koppelten Prozesses von Zuckerverbrennung und Glykogensyn-
these».

Staub (1930), summing up, states; ^Der normalphysiologische
Insulineffekt am gesunden, glykogenarmen Organismus besteht
demnach in Beschleunigung der Glykogenbildung im ganzen Orga-
nismus und wahrscheinlich auch in der Leber. Mittlere toxische
Dosen vermindem nur den Glykogengehalt der Leber and reichern
denjenigen des iibrigen Organismus etAvas an; es kommt aber im
ganzen zu einem Glykogendefizit. Hochtoxische Insulingaben ver-
ringern sowohl Leber- wie Muskelglykogengehalt erheblichi.

C. F. CoRl (1928) considers that a marked increase of oxidation
and storage of sugar is found in the peripheral tissues after the

injection of insulin, but that the liver glycogen content is greatly
reduced. And this occurs even if the animal has carbohydrates
available in its intestine in the meantime. He comes to the foll-
owing conclusion: i.Tliis discussion shows that our information con-
cerning the influence of insulin on the carbohydrate metabolism
of the liver is still very scanty. There are no convincing exper-
iments on record which would show a direct influence of this hor-
mone on either glycogen synthesis or glycogenolysis. The exper-
iments on the perfused liver of mammals and of cold-blooded ani-
mals did not yield results that permit of a definite interpretation.
In normal fasting animals the majority of the investigators observ-
ed a decrease in liver glycogen after the insulin injection . . .»

It appears from the foregoing discussion that further work is
needed to elucidate the relation between insulin and the carbo-
hydrate metabolism of the liver.

Geiling, Jensen and Farras (1937), summing up, state: »In
contrast to its action upon the diabetic, insulin usually decreases
the glycogen stores of the normal fed and starved animal. Its
effect upon the liver and other glycogen depots is, however, a
complex matter, difficult to interpret in the light of our present
imperfect knowledge concerning the mechanism of insulin action* .

Lundsgaard (1939) writes: *Today, just as was the case im-
mediately after the discovery of insulin, the question of how in-
sulin works can be answered only by reference to the directly ob-
servable effect of insulin in the animal organism; a reduction of
the blood sugar concentration. If one is asked as to what causes
this reduction of the blood sugar concentration, the only positive
answer is that the effect, partly at any rate, is the result of an
acceleration of the rate of migration of glucose into the striated
muscle fibres from the blood and an interconnected acceleration of
the glycogen deposition in the muscles. It is doubtful if insulin
has a corresponding effect in the liver.*

Gerhitzen (1942) states it to be the general view in modern
times that » insulin promotes the fixation of glycogen in the liver,
or, perhaps better expressed: checks the flow of glucose into the
blood, thus giving rise to a larger content of glycogen in the liver.
Insulin promotes the deposition of glycogen in the muscle.»

Best and Taylor (1943) state: i-In the normal animal one of
of the most clear-cut effects of insulin is the increase in the rate
of glycogen deposition in the muscle which it produces, but it has
not been demonstrated in any normal adult animal that insulin
increases the level of liver glycogen. The increase in the livers of
young rats is due to a secondary liberation of adrenaline. In the
normal adult animal there may be an actual loss of glycogen from
the liver when insulin is administered. This is due to the acceler-
ated glycogen deposition in muscle and the increased rate of oxid-
ation of sugar.j-

A systematic review of the immense number of investigations
published on this question falls beyond the scope of this work.
Here it must suffice to mention some of the works most cited in
the literature as well as those which have an important bearing
on the discussion.

The present investigation is intended to obtain reliable data re-
garding the effect of insulin on the glycogen stores of the body,
when it is supplied in excess to normal animals. This would provide
a firm basis for continued studies on the effect of insulin.

The investigation falls into two parts: 1) Study of the conditions
for comparative glycogen investigations, i. e. a study of the factors
affecting the glycogen storage in normal animals and an inquiry
as to how the least possible variation in these values can be
attained. 2) Study on the effect of insulin on the glycogen reserves
of the body in normal animals.

PART I

METHODS

CHAPTER .1.

The glycogen analysis method

A. Description of the method.

The glycogen determinations in the present investigation were
made in accordance with the modification of Pfluger’s method
indicated by Sjogren, Nordenskjold, Holmgren and Mol-
lerstrom (1938).

The analysis is made in the following manner:

I. Ldberation of the glycogen: Centrifugal tubes are charged
with 2 — 3 cc 30 % KOH and are heated in a boiling water bath.
Suitable pieces of the organs are weighed rapidly on a torsion
balance with a sensitivity of 10 mg and are plunged into the hot
potash lye. If large pieces of tissue, over 1 gram, are taken, the
amount of potash lye must be increased in proportion. Boiling for
30 minutes, with frequent shaking.

II. Precipitation and washing of the glycogen: After cooling of
the tube, 0.5 cc of a saturated sodium sulphate solution is added
as well as spiritus concentratus in sufficient amount to bring the
grade of alcohol up to 75 — 80 %. Afterwards heating in a water
bath to boiling point, cooling in running water and centrifuging.
The clear motherdye is cautiously poured away. The precipitate
IS dissolved in a small amount of distilled water (2 — 3 cc) and the
precipitation with sodium sulphate, alcohol and boiling is repeat-
ed. Cooling, centrifuging and decantering. The precipitate is again
dissolved in Avater and precipitation is once more made, now with-
out the addition of sodium sulphate, with alcohol solely.

HI. Hydrolysis. The precipitate is dissolved in 3 cc of Avater
3nd exactly 3 cc of 10 N sulphuric acid is added. Hydrolysis is
effected by boiling in a Avater bath with an air-cooler for 30
minutes.

IV. NeutraUzatUyti and dilution: After hydrolysis the solution
is neutralized with 3 N NaOH, with phenolphthalein as an indicat-
or. The sample is diluted in a volumetric flask precisely to the
required volume.

V. Determination of the amount of glucose in the diluted
sample: Exactly 5 cc of the solution is mixed in a pyrex tube with
exactly 5 cc of ScHAFitER-SoMoeyi’s reagent. The tube is plung-
ed into a boiling water bath,' is allowed to boil for precisely 15
minutes, and is then quickly cooled in cold running water. There-
upon 2 cc of a solution containing 2.5 % of potassium iodide and
2.5 % of sodium oxalate is added, and then 5 cc N HgSO^. The
sample is shaken now and then, and after 10 minutes is ready for
titration with ^^7200 sodium thiosulphate, with starch as an
indicator.

The analysis sample, after titration, should contain 0.5 — ^2.o mg
glucose. If it contains more, the method will give too low values,
and the analysis will have to be repeated with a smaller amount
of test-solution (4 — 3 — 2 — 1 cc) and, instead, addition of 1 — 2 —
— 3 — 4 cc distilled water. Should too high glucose values be ob-
tained even with 1 cc test-solution, a new dilution must be made.

The amount of glucose is found by readings on a comparison
curve, which is obtained by measuring in four tubes 2, 3, 4 and 5
ml of a freshly prepared solution of exactly 50 mg pure glucose per
100 cc, adding 3, 2, 1 and 0 cc of water, so that all the tubes con-
tain 5 ml, and then 5 cc Schaffer-Somogyi’s reagent to each
tube; analysis as above. The amounts of thiosulphate obtained
are marked on a curve with the amount of glucose in the test-
solution along the x-axis and the amount of thiosulphate along
the y-axis. At least three of the values obtained should lie along
a straight line.

The value read off on the curve indicates the amount of glucose
in the analysis test. From it the amount of glucose in the organ
sample is computed. The glycogen value is obtained by multipli-
cation with the factor 0.927 (Pfluger).

According to Sjogren, Nordenskjold, Holmgren and Mol-
lebstrom, this method gives results with an accuracy of ± 5 %.
». . . welche als die zur Zeit optimale anzusehen ist>. Some
earlier authors have spoken of analyses with a precision of ± 2 %■

This, however, is considered by Sjoghen and his co-workers to
be a considerable overestimate of the exactitude of the method.
As for the yield, it is not mentioned at all by Sjogren et alios.
For a similar method, Sahtdn (1931) reports a yield of about
100 % on analysis of glycogen amounts ranging between 5.o and
160 mg. His series, however, are very small, and we cannot gather
from his published report whether the figures are based on the
results of individual analyses, or whether they are averages for
serial analyses. He gives no figures for the deviation. — Similar fi-
gures are reported by other authors.

It seems a ‘priori improbable that a chemical method of this na-
ture should give a yield of 100 per cent. Evidently, none of the
various processes involved in the method could be carried on en-
tirely without loss, though the loss may be very small. In the
present investigation therefore the accuracy of the method has
been tested with a view to ascertaining (1) the percentage of the
yield, (2) where in the course of the analysis a possible loss occurs
and (3) the constancy of the results, which, of course, is a matter
of special importance in practice.

B. The testing of the method was made in the
following manner.

1. Testing of the glucose analysis method. As the glucose de-
termination is based on the reading of a comparison curve, and
as that curve is traced from four points by analysis of four samples
containing known amounts of glycogen, it is essential to ascer-
tain whether the straight line thus obtained can actually be applied
to higher and lower glucose values than those used in the construc-
tion of the curve. Another important matter is the determination
of the accuracy of the individual analysis. These two factors were
determined by serial analyses of pure glucose. Different amounts
of glucose solutions with a known concentration were analyzed.
A curve was drawn on the basis of the average consumption of
thiosulphate for the amounts of glucose ordinarily used in the con-
struction of such curves. This curve was extended in both direc-
tions. The values for the consumption of thiosulphate with the

Fig. 1. Diagram showing required, volume of Na^SJ)^ when hnown
quantities of -pure glucose are determined.

curve drawn as described in the text.

curve based on estimations including excessively small or

large quantities of glucose.

other analyzed amounts of glucose were then inserted. See the
diagram Fig. 1 and the table Fig. 2.

Figs. 1 and 2 show that the curve for the consumption of thio-
.sulphate is linear as regards glucose amounts ranging between 0.5
and 2.5 mg. Whether the' amounts of glucose in the samples analyz-
ed are larger or smaller than these limit values, relatively larger
quantities of thiosulphate will be consumed. Thus, if a comparison
curve is traced in the usual way, the glucose readings will be too
low. As will be seen from the table in Fig, 2, the standard de-
viation of the consumption of thiosulphate in fact varies in re-
spect of different amounts of glucose. It shows a marked increase
when the amount of glucose reaches 2.5 mg or more. However, if
we glance at the column showing the deviation expressed in glu-
cose, it will be seen that the variation in the result of the analysis
keeps fairly constant, though it rises somewhat for the larger
amounts of glucose. If, on the other hand, this deviation is placed
in relation to the amount of glucose in the sample, i. e. if we look

Fig. 2. Table. Determinations of known quantities of pure glucose. Control
of the variability of the method of glucose estimation.

Quantity of glucose
added. Mgm.

Number of determinations

Eequired vol. of
Na 2 S 203

Standard deviation of required vol.
of NaoSaOs given as corresponding
quantity of glucose according to

the correct
curve. Fig. 1

the constructed
curve. Fig. 1

Standard

deviation

Glucose

mgm

Per cent
of glucose
added

Glucose

mgm

Per cent
of glucose
added

0.1

22.5

0.08

0.2

21.7

■■

O.io

0.3

21.1

0.6

O.IO

0.4

20.8

0.5

O.08

0.6

19.7

0.6

—

0.10

1.0

16.8

0.6

—

1.5

13.9

0.6

—

0.10

2.0

11.2

0.6

—

0.10

2.5

8.2

0.8

—

0.18

3.1

5.1

1.2

—

0.21

3.6

3.8

0.8

—

0.18

at the standard deviation in per cent of the mean, it will be
found that it is much larger for small amounts of glucose, 0.5 — 0.4
mg and under. As regards other analyzed amounts of glucose, it
is fairly constant.

As is clearly indicated by these results, the analyses should be
so adjusted that the amounts of glucose in the samples analyzed
range between 0.5 and 2.5 mg. This in fact was the rule adopted
throughout the present investigation, with the exception of a few
series where, owing to the low glycogen content in the organs,
it was not found possible to reach such high values even when the
whole organ was analyzed. However, a glycogen value below 0.4 mg
per analyzed sample hardly ever occurred. These values too were
read on the ordinary comparison curve. In fasting mice, which
have a very low glycogen content, this was partly compensated
by the analysis of the livers in couples.

Fig. 3. Table showing the result of repeated glucose determinations of one
and the same dilution of a given liver glycogen sample. The sample was
always diluted to such a degree that the required volume Na„S.X)^ should
fall within the optimal part of the test curve.

Number of
determina-
tions

Liver glyco-
gen content
per cent

Hange

Standard

deviation

Standard
deviation in
per cent
of mean

■n

3.18

3.48— 2.97

O.IG

b.O

8.74

9.27— 8.14

0.39

4.6

10.96

12.00— 9.93

0.77

7.0

11.74

12.46—10.82

0.65

4.7

The exactitude of the sugar analysis method was also determ-
ined by analyses of liver samples. When the sample had passed
through the whole analysis procedure including the dilution, a
series of sugar determinations was made. The results are tabulated
in Fig, 3, which gives a good idea of the accuracy of the method
in practical work with different amounts of glucose. The standard
deviation in per cent of the mean for all the values is fairly
constant.

The sugar analysis method was controlled in yet another way,
namely by double analyses of liver and muscle glycogen samples
from mice. As will be seen from the table in Fig. 4, the values
show throughout a very good correspondence. Nowhere is the
mean difference statistically significant, that is to say, there is no

Fig. If., Table. Double glucose determinations of liver and
body glycogen of mice.

Organ

No of
determi-
nations

Sample I
Glycogen
content
per cent

Sample 11
Glycogen
content
per cent

Mean difference
and its standard
error

dies

Liver I

1.08

1.06

— 0.08 ±0.08

Liver n . . . .

3.38

0 ± O.Ol

Muscle I . . . .

0.046

0.044 1

— 0.002 ±0.001

Muscle n . . .

0.1 40

0.144

0.004 ±0.006

Fig. 5 . Table

Control of method. Determinations of known quantities of
■pure glycogen. Direct hydrolysis of glycogen.

Glycogen

added

mgm

No. of
samples

Recovered

glycogen

mgm

Standard

deviation

Standard
deviation in
per cent of
mean

Yield
per cent

1.86

1.26

0.18

10.4

3.72

3.12

0.14

4.6

84:

5.68

5.18

0.81

6.0

16.74

15.26

0.82

5.4

33.48

30.07

1.48

4.9

34.88

31.00

2.28

7.2

systematic difference between the values obtained in the two
analyses.

These three control tests thus show that the glucose analysis
method gives very reliable readings if the samples analyzed
are adjusted so as to contain 0.5 — ^2,o mg glucose. Should the
amounts of glucose be larger or smaller, too low glucose values
will be obtained if a comparison curve is traced in the above
stated manner.

Besides the sugar analysis, there are two other factors in the
glycogen determination that may conceivably entail losses,
hydrolysis and the washing process.

2. The effect of hydrolysis cm the yield. In these tests, known
amounts of pure glycogen per analysis were hydrolyzed and e
termined. All the sugar analyses were made in such a \yay that
the consumption of thiosulphate fell, so far as possible, within t e
optimal part of the comparison curve. The glycogen used had a
water content of 7.oi % and an ash content of O.0029 In t e
table Pig. 5 and in the diagrams Figs. 6 and 7, the amounts of
supplied glycogen refer to the >solids». As shown both by the ta
and by the diagrams, the results of the analysis are very uniforin.
Apart from the lowest amounts of glycogen, the standard devi
ation in per cent of the mean is constant. The losses are
throughout quite small; they will, of course be relatively larger

Fig. 6. Diagram, showing yield, when known quantities of ‘pure
glycogen are determined.

Percent

C ZO ftO BG Mgm

the sample was immediately hydrolyzed.

the sample passed through all the procedures of the method.

including precipitation and washing.'

Fig. 7. Diagram showing yield when pure glycogen is determined.

if very small amounts of glycogen are analyzed, in which case the
percentage yield will be poorer.

3. The effect of the washing and precipitation processes on the
yield. The results of the analyses of pure glycogen which has
passed through the whole course of analysis are tabulated in Fig. 8
and are graphically shown in the diagrams Figs. 6 and 7. The
yield will be somewhat poorer than in the direct hydrolysis of

Ftg. 8. Table. Control of method. Determination of pure glycogen. The
samples passes through all the procedures of the method, including
usual washing and precipitation before hydrolysis.

Glycogen

No. of

Glycogen

Standard

Standard
deviation in

added

determi-

recovered

deviation

per cent of

Yield

mgm

nations

the mean

per cent

1.8G

1.09

0.21

19.3

3.72

2.81

O.60

17.8

5.58

4.39

0.33

7.6

11.16

9.26

0.66

7.0

22.32

18.97

1.28

6.7

44.64

37.73

2.10

5.6

the glycogen, Avhich indicates a small loss in the precipitation
and washing procedure. This agrees well with the investigations

mg of glycogen is dissolved
la cc 80 % alcohol. The standard deviation in per cent of
he mean is also somewhat higher than in direct hydrolysis: but,
'V en the amount of glycogen is not less than 5 mg, it remains
constant, as also the percentage yield. The latter is, of course,
poorer for low amounts of glycogen.

d ^^f'^^ceableness of the method is also indicated by the
mi^ cterminations of the glycogen content in liver and
oil p 53 tabulated in Figs. 20 and 21 and are discussed

Work control tests show that the method, if we do not

adh' small amounts of glycogen, and if the analysis is
the 0 V 3^niount of thiosulphate consumed falls within

fact comparison curve, operates with satis-

cnsueT gives an acceptable yield. A minor loss

Joss Tjf analysis, but nowhere any considerable

anaiysis^^^ particularly weak spot in the course of the

dire noted that all the glycogen values are figures

° tained in the analysis and have not been corrected for
bosses during the course of it.

C. other reducing substances than glucose possibly
found in the hydrolysate.

In the analysis method it is, of course, intended that all other
substances than glycogen should be broken down and washed
away, so that after the washing only glycogen is left in the sample
analyzed. Consequently, glucose should be the only reducing
substance in the sample after hydrolysis.

CoRI (1928) states: i-The non-sugar reducing substances of the
hydrolysate were determined after removal of the sugar with
copper sulphate and lime. Since they constituted only 4 to 5 per
cent of the total reduction . . .» CORI and CoRl (1930), however,
after examining muscles in which they had determined the
hydrolyzed glycogen before and after fermentation of the sample,
report that sthe reducing power in terms of glucose after fermenta-
tion of the glycogen hydrolysate was between 0 and 12 mg. per
100 gm of muscle. This small correction of the glycogen values
could safely have been omitted without changing the significance
of the results.*

Sjogren, Nordenskjold, Holmgren and Mollerstrom
(1938), on examining analyzed samples of liver, found that they
contained a reducing substance which they designate as >non-
sugar*. It is considered by them to represent 10 per cent of the
total reduction, and this figure is stated to be fairly constant.
Sato (1923) considered he had found a *Ilestreduktion, die durcli
nicht-glykogene Substanzen verursacht worden ist*. It is, he
states, fairly constant in amount and *er betragt bei Kaninchen
in der Leber 0.3 % (als Glukose berechnet), in den Muskeln
0.2 %», and is independent of the content of glycogen. Sato,
however, worked with another glycogen determination method,
a modification of that of Bierry-Gruzewska, in which the gly-
cogen is hydrolyzed without washing. His figures therefore are
not applicable to the results of the Peluger method.

May (1934, a) isolated from a snail a new polysaccharide of galac-
tose, which he terms *galaktogen*. In another paper (1934, b),
he shows that ^galaktogen* occurs also in the mammalian orga-
nism under certain conditions in rather large amounts, and that

it is included also in analyses of glycogen. May and Wein-
BREIWER (1938) consider that j>galaktogen» is a normal consti-
tuent in the mammalian organism, but that in adult males and
non-lactating females it occurs merely in insignificant amounts.

As, in my experiments, I always work with adult, non-lactating
animals, any »galaktogen» that may possibly occur in connection
with them would seem to be a quantite negligeable. According to
the author’s above reported investigations, any 3>non-sugar»
reducing substances that may possible occur must be present in
such small amounts that they do not affect the results and no
regard has therefore been paid to them in the present investiga-
tion.

CHAPTER 2.

Statistical metliods

Symbols:

X Mean

a Standard deviation

e Standard error of a mean = —

i/n

d Difference
d Mean difference
n Number of variates

T«i Standard deviation in per cent of the mean.

Methods :

I. For computation of the significance of the differences
between two means, derived from relatively small groups of ani-
mals, t-analysis (iStudentj 1908, Fisher 1936, Bonnier and
Teddin 1940) was adopted.

Vex^ + fiy-

df degrees of freedom (in t-analysis = n^^ n 2 — 2)

P Probability that the groups coincide.

P 0.05 = coincidence, corresponding to a difference of less
than 2 a in large groups.

P O.05 — 0.01 = a probable difference, corresponding to a
difference of 2 — 2.5 a.

P O.oi — 0.003 =: a highly probable difference, corresponding
to a difference of 2.5 — 3 o.

P O.003 = a significant difference, corresponding to a differ-
ence of more than 3 a.

n. For computation of the significance of differences between
the means derived from more than two groups of samples, ana-
lysis of variance (Fischer 1936, Bonnier and Teddin 1940)
was adopted, the variance ratio (F) being calculated according to
Snedecor (1940).

CHAPTER 3.

Eoutine of tlie experiments

As already pointed out, and as will be further explained in the
sequel (p. 43), even minor deviations in the experimental con-
ditions may entail marked variations in the amount of the body
glycogen deposited. It is thus of the greatest importance
that the experimental conditions should, so far as possible, be
standardized. I therefore give here a brief review of the general
routine of the experiments. Further particulars will be reported
in the next section of this work.

Except where otherwise stated, the animals were treated in the
following way; —

Uniform breed. As has been frequently pointed out, it is of the
greatest importance that the experimental animals should be of
the same breed. In fact, many investigators have taken their
laboratory animals and controls from the same litter. Orten and
Sayers (1942) state that they had found a distinct difference
between two different breeds of rats. — Owing to lack of accom-
modation, the author has found it impossible to use a uniform
breed throughout, but he has tried to secure the greatest possible
uniformity by purchasing almost all the animals from the same
breeder. At any rate, all the animals in a series and the corre-
sponding controls came from the same supplier.

The animals were left at least one week in the ^stablei^, under
uniform conditions, before they were used for experiments, in order
to compensate possible differences due to transport and change
of environment.

The diet for mice consisted of ordinary bread. Card’s »mouse
bread*, as well as milk and water ad libitum. The rats were
put on the same diet. The rabbits were fed on hay, oats and

turnips. During the summer, M'hen merely a few experiments
were made, the turnips were replaced by fresh green grass.

On these diets the animals grow normally. They are sprightly
and lively, show no signs of deficiency, and the fertility is normal.

Composition of the »mouse breads : 100 kg
of milk, 53 kg of crumbs (unspiced), 20 kg of oaten groats, 15 kg
of wheat sprouts, 10 kg of lucerne meal, 2.5 kg of minced meat
and 40 gm of cod liver oil.

The temferature of the environment could not be kept quite
constant, but varied between 16° and 21° C. during the periods
when the reported experiments were made.

As the age of the animals may affect the amount of the glycogen
depots, animals of about the same age were always used, their
weight being taken as an indication of their age.

Sex. Almost all the mice used were female. As regards rats and
rabbits, it was found impossible to procure a sufficient number of
one sex: for this reason, each series included as a rule an equal
number of males and females.

Controls: The controls were always examined concurrently with
the experimental animals.

Mode of killing: Apart from certain special series, all the animals
were killed by decapitation with a heavy axe. Immediately after-
wards, they received a blow which broke the lumbar cord. The
postmortal spasms in the body were thus avoided.

Dissection for analysis, etc: As soon as possible after death, a
slice of liver is dissected for analysis, is weighed on a torsion
balance Avith a scale of 10 mg, and is plunged into hot KOH
within 50 seconds after decapitation. As regards rabbits and rats,
a piece of muscle is dissected, weighed and immersed in hot KOH
within 90 seconds after decapitation. In rabbits m. triceps sin.
IS regularly used, in rats m. triceps surae and plantaris from one
side or both. In mice, the liver is fitst dissected, weighed and laid
m hot KOH. Then the spleen and alimentary canal from the
stomach to the rectum are removed. The rest of the animal is cut
into four pieces, which are plunged into hot KOH within 90
seconds after decapitation.

Serial experiments: All the experiments were serial, and all com-
parisons are based on the averages from the series.

Insulin: A crystalline insulin of the Vitrum brand was used
throughout. It contains 22 international units (U) per mg and is
dissolved in physiological saline solution. The dosage is always
given in U.

Injection: The rabbits remain quite quiet during injection. The
rats are held with a firm grip over the lower jaw, the body being
otherwise free. The mice are left free on the table and the insulin
is injected subcutaneously, whilst they are pulled lightly by the
tail.

Care is taken not to excite the animals, and they are always
treated as gently as possible.

PART TI

SURVEY OF FACTORS
NORMALLY INFLUENCING THE
GLYCOGEN STORAGE

CHAPTER 4.

The glycogen reserves of the body

The important substance glycogen was first detected and
studied by Claude Bernard. He discovered in 1850 that the
liver gave off sugar to the blood. He considered at first that
this sugar developed from protein, but in a series of works in the
course of the immediately following years he indicated the
principal properties of glycogen; in 1855 he presumed that it was
a jsorte fecule animale» in the liver, and in 1857 he suggested
the method for its production in a pure form. Glycogen has since
been of central importance for the study of the metabolism.
Claude Bernard had already indicated the liver and the skeletal
muscles as the principal storehouses of the body for glycogen.
This view has subsequently been confirmed and accepted, and
investigations of variations in the glycogen reserves of the body
have, generally speaking, been confined mainly to those two
organs. Other organs contain merely minor amounts of glycogen.

Under certain special conditions, however, the adipose tissue
may store considerable amounts of glycogen and sometimes it
may actually have a glycogen content otherwise found only in
the liver. In normal animals the adipose tissue as a rule contains
merely insignificant amounts of glycogen. Gierke (1907) showed
that the adipose tissue in guinea-pigs which, after fasting for
three days, had been put on abundant diet, was still free from
glycogen after 2 — 3 days, but after 7 days contained considerable
amounts of glycogen, Avhich then again decreased. The fact that/
the adipose tissue in animals which, after fasting for some length
of time, had been put on a high carbohydrate diet, during a
certain period contained an abundance of glycogen has afterwards
been confirmed, for example by Arndx (1927) on dogs, rabbits
and man, Hoffmann and Wertheimer (1927) on dogs, Wert-

HEiMER (1928), Eichter (1931), Eger and Morgenstern (1938)
and Eger (1942) on rats. That considerable amounts of glycogen
may occur in fat is indicated by the fact that Arndt, under these
conditions, found a glycogen content of up to 7.4 %, and Eger
up to 6 %, in the adipose tissue. Richter moreover ascertained
that the administration of insulin had no effect on this storage
of glycogen in the adipose tissue. According to Scoz (1929), con-
siderable amounts of glycogen are stored in the subcutaneous
tissue of dogs under similar conditions.

Glycogen thus occurs in large amounts in the adipose tissue
only under certain special conditions, lengthy fasting and after-
wards an abundant diet for some length of time. These conditions
were not provided in my experiments, and my chief interest was
therefore devoted to the study of the glycogen variations in liver
and muscles. Only in a couple of smaller series on rats was the
content of glycogen in the skin also analyzed. In studies on
mice the whole body was analyzed, whence the glycogen varia-
tions in the adipose tissue are included in the total.

It is generally known that the glycogen reserves are subject
to considerable variations. Ppluger, as far back as 1902, clearly
realized the very marked individual fluctuations, and in recent
times attempts have been made to standardize the experimental
conditions, in order as far as possible to work with glycogen
amounts of the same magnitude. The requirements that are
nowadays set up have been clearly formulated by Guest and
Rawson (1939); j>In glycogen determinations large deviations
from animal to animal appear to be characteristic. In order to
materially reduce this variability a standardization of controls has
been attempted. The precautions taken and the results obtained
are outlined below:

1) An inbred strain of Wistar rats is used.

2) Only males are used.

3) The age of the animals is restricted to 100 ± 4 days at the
time of the sampling.

4) A uniform dry pellet diet wh’ch contains more than the
minimum of all factors necessary for the maintenance of growth
and health is fed.

5) An inanition period is established during the interval from
72 to 12 hours preceding the sampling. Food is given ad lib.
during the final 12 hours, beginning at 10 p. m.

6) The liver and muscle samples arc taken between 9:30 and
11: 30 a. m.

7) The weight of the dried stomach contents is subtracted from
the weight of all the food eaten and all animals in which less
than 35 mgm. per gram of body weight has passed the stomach
are e.\'cluded.

8) Since rats are noclurnally active, a light is left on until
10 p. m. at the lime the feeding begins.

9) The temperature of the environment is maintained at 28° C.
during the 72 hours preceding the sampling.

10) Anesthesia is bj’ intraperiloncal injection of O.'o mgm, per
kgm body weight, of nembutal. Induction occurs within 4
minutes.*

The question as to the effect which all these different factors
may have on the amount of the glycogen reserves is, of course,
of the greatest importance for an investigation of this nature.
I will therefore give a brief review of the various factors which arc
considered to affect the amount of the gljmogen reserves in normal
animals.

CHAPTER 5.

Distribution of glycogen in tlie
glycogen reserves

The distribution of glycogen within the reserves is a problem
of the greatest importance if we work with micromethods and let
a small piece, or a few pieces, of an organ represent the organ in
its entirety. Such procedure is, of course, permissible only if the
distribution of glycogen is uniform over the whole organ. If the
entire organ is subjected to analysis, this problem will, of course,
be eliminated.

A. The glycogen distribution in the liver has been the subject
of much discussion. Thus, for example, Lttchsinger (1875),
Seegen and Kratschmer (1880), Kulz (1886) and Cramer
(1888) considered that the distribution was uniform over the
whole liver, whereas v. Wittich (1875), Barfurth (1885) and
other investigators contended that it varied. Grube (1905) con-
sidered that the glycogen content was uniform throughout in the
actual hepatic parenchyma, but that it varied in different parts
of the liver in correspondence with the amount of connective
tissue. Among later investigators, Schondorfp (1903), Maceeod
and Pearce (1911), Paulesco (1913), Folin, Trimble and
Ne^vman (1927), Evans, Tsai and Young (1931), Holmgren
(1936) and Sjogren, Nordenskjold, Holmgren and Moller-
strom (1938) support the view that the distribution is uniform,
whereas e. g. Scheiff (1931) considers that it varies. This ques-
tion is discussed — after a review of the literature and in relation
to the theories regarding the distribution of the portal blood
stream in different parts of the liver — by Henschen (1932),
Holmgren (1936) and other authors.

Fig. 9. Table. Comparison between glycogen content of right and left part
of the liver in normal rats. Sample of right part of the liver in hot KOH
within 50 secs, of decapitation and sample
of left part within 90 secs.

No. of
experiments

Glycogen
content of right
part

per cent

Glycogen
content of left
part
per cent

Mean difference
and its

standard error
diej

5.G9

5.68

~ O.Ol ±0.18

Own investigations. In view of the great importance of this
problem and the disparity of the results, I have studied this ques-
tion in experiments on rabbits and rats. As regards rats, the
analyses were made in the following manner: — As soon as the
untreated animals had been decapitated, the abdomen Avas opened,
and a piece of the right part of the liver was cut off, weighed
and plunged into hot KOH, within 50 seconds after decapitation.
Immediately afterwards this procedure was repeated with a piece
of the left part of the liver and this piece was plunged into hot KOH
within 90 seconds after decapitation. According to my investiga-
tions, the postmortal glycogenolysis has not yet set in so early
after decapitation (cf pp. 4-7 and 53).

The results of my study of the glycogen distribution in the liver
of rats are tabulated in Fig. 9. These investigations show that
there is no systematic difference in glycogen content between the
two halves of the rat liver.

In my studies of the rabbit liver the procedure was as follows: —

Prom the right part of the liver a slice of liver Avas cut
out and divided into a number of pieces of suitable size. They
Avere Aveighed as rapidly as possible and plunged into hot
KOH. This procedure Avas then repeated AA-ith a similar piece
from the left part of the liver. All the pieces were placed in hot
KOH Avithin 5 minutes after decapitation. According to my in-
vestigations, the postmortal glycogenolysis in the liver has indeed
begun 5 minutes after decapitation, but it has not then yet attain-
ed such magnitude that the liver glycogen values statistically

Fig. 10. Table showing the glycogen -percentage of different parts of the liver in normal rabbits. The abdominal wall
was cut open immediately after decapitation. A long rodshaped piece from the right liver lobe was cut into smaller
pieces, which were laid in hot KOH. Immediately afterwards the same procedure was repeated with a piece of ike left

o a

o a a

a a Ci

s s a

Sc®

kO CO

O O

O O O

+ 1 + 1+1

A lO O

o o d

1 !

o o o

+ 1 + 1+1

A O C*

oo-

d CO d

O lO X CO CO

o o o o o

+! +1 +1 +1 +1

o oo CO

O CO A 1-.

CO od* T-< 00 d

CD CO CO CD

<j PQ a o w

2.lf,±G.4:

difiEer from those obtained on analysis within 50 seconds after
decapitation.

Five rabbit livers were analyzed in this way, and the results
are tabulated in Fig. 10. As indicated by the t-analysis, a differ-
ence between the glycogen content of the two hepatic lobes
appeared to be probable in one case. If, however, the mean differ-
ence is computed, no systematic difference will be found. The
results do not rule out the possibility of some difference in gly-
cogen content between the hepatic lobes, but they argue against
the systematic occurrence of such a difference. That the post-
mortal glycogenolysis can have no essential bearing on these results
is indicated by the fact that sometimes more, sometimes less gly-
cogen is found in the subsequently analyzed piece.

If pieces of liver are taken at random from a series of animals,
a serviceable average value should thus be obtained. That this
IS in fact the case is shown by the analyses of rabbit liver made
in connection with the study of the initial postmortal glycogen-
olysis (see p. 50). They were made on pieces of liver taken at
random from different parts, at as short intervals as possible.
The results are tabulated in Fig. 19. The table shows that, when
liver samples are taken serially from animals, comparable glyco-
gen values will be obtained, which in this connection is the
essential.

B. The distribution of glycogen in the skeletal muscles. Also
as regards the skeletal muscles, we must as a rule content our-
selves Avith analyzing a small sample and letting it represent the
whole musculature. Only small animals, such as mice and possibly
rats, can be studied and analyzed in to to. It is therefore of funda-
niental importance to ascertain whether the different muscles of
the body have the same glycogen content, or whether there are
regular variations between different muscles.

As regards mammals, Nasse (1877) considers himself to have
found considerable variations, likewise Cramer (1888), Moscati
(1907) and Choi (1928). On the other hand, Elias and Schu-
bert (1918), Folin, Trimble and Newman (1927), Long
(1928) as well as Evans, Tsai and Young (1931) find good corre-
spondence between different muscles. Best, Hoet and Marks

(1926) likewise find a good correspondence between the sym-
metrical muscles in cats. Sahtdn, Simmonds and Working
(1934i), whilst finding a good correspondence between the sym-
metrical muscles of rats, state that they had observed a higher
glycogen content in the muscles of the hindlegs than in those
of the fore legs. Masatma and Riesser (1931) find a higher
glycogen content in the Avhite than in the red muscles in rabbits.

A number of studies on the glycogen content in different muscles
of birds have been published, but seem to be devoid of interest
in this connection.

Several investigators, such as Corkill (1930) and Goldblatt
(1930), have simply taken a single muscle as representative of
the whole skeletal musculature.

The author’s own investigations were made on rabbit muscles.
They were firstly in the nature of double determinations of the
glycogen content in different parts of the same muscle, from
which two pieces were taken for analysis as soon as possible
after decapitation. — This material will be discussed in connec-
tion with the question of the initial postmortal glycogenolysis,
and the results are tabulated in Fig. 20.

Secondly, in a series of normal rabbits, different muscles were
dissected as soon as possible after decapitation and analyzed for
their glycogen content. The first muscle was plunged into hot
KOH within 50 seconds after decapitation and the last within
5 minutes. As shown by the studies on postmortal glycogen-
olysis reported further on, no change in the glycogen content of
the muscles, as compared with the sample taken after the lapse
of 50 seconds, occurs during the first 15 minutes after decapita-
tion.

The results of these investigations are tabulated in Fig. H-
This table shows that there is no difference in glycogen content
in the different muscles examined. In Fig. 12 the same values are
tabulated in the form of averages for the glycogen content of
different muscles in the individual animals. Here we find a marked
variation and apparently in several cases statistically significant
differences. If this material, however, is subjected to an analysis

Fk 11. Table. Estimation oj glycogen content of different mvsclcs in the
rabbit. The mnscles were itrepared as rapidly as posmble and laid in hot
KOH within 50 secs, to 5 mins, of decapitation. The muscles were
prepared in the order given in the table.

Muscle

No. of
examined
muscles

Glycogen content.
per cent

0.25 ±0.03

0.28 + 0.04

0.2G ±0.03

0.31 ±0.03

0.36 ±0.05

M. semitendineus (red)

0.38 ±0.0G

of variance (see Fig. 13), it will be found that these variations are
due to chance.

Thus, in studies of this nature we seem to be warranted in
reckoning with a uniform distribution of the glycogen not only in
the same muscle, but also within the entire skeletal musculature,
so that a glycogen value from a single muscle may be considered
to be representative of the whole system.

Fig. 12. Table. Results shown in Fig. 11 grouped together for every separate
animal and given as the average glycogen content of its muscles. TF lere
only five muscles were examined, m. semitendineus
was always the one omitted.

No. of
muscles

Glycogen content
per cent

X±£-

No. of
muscles

Glycogen content
per cent

X±£5j

0.21 + O.Ol

0.37 ± 0.04

0.26 + 0.02

0.40 ±0.03

0.55 + O.04

0.31 ±0.01

0.30 + 0.04

0.13 ±0.02

0.36 + 0.04

0.19 ±0.04

0.27 ±0.03

0.16 ±0.01

0.24 ±0.04

Fig. 13. Table showing the resnlts of an analysis of variance of the
res^dts tabulated in Fig. 12.

Cause of variation

Degrees

freedom

Square

Mean square

Between the groups ....

0.8531

0.0711

Within the groups

0.34)8

0.0068

Total

1.1949

p,_0.0711_

0.0058

1.226

P = > 0.2

Summary:

No systematic difference in the distribution of glycogen in
different parts of the glycogen reserves could be shown. Even
if rather marked differences between different parts of the same
animal may occur, it will doubtless be possible, by analyzing
part of the glycogen resen^e from series of animals, to obtain
an average value which is representative of the reserve as a whole.
This applies both to the liver and to the skeletal muscles.

CHAPTER G.

Mode of killing .and postmortal
glycogeiiolysis

I. Mode of killing. Generally speaking, two methods of
killing the animals are reported in the literature: Decapitation
and narcosis.

SAHYnN" and Luck (1929), in their glycogen analyses, take
pieces of muscle from the hind legs of the animals, The,y consider
that the sciatic nerve should be cut off, in order to avoid spasms
and a consequent loss of glycogen. Anderson and Macleod
(1930) state that the postmortal spasms are largely accountable
for the individual differences in the content of muscle glycogen.
Macleod and Pearce (1911) state that these spasms can be
avoided by decapitating the animals in the upper thoracic region.

Evans, Tsai and Young (1931) consider that more liver
glycogen is lost if the animals are decapitated than if they are
narcotized with ether. Corn (1931), however, recommends large
doses of amytal intravenously for cats and rabbits, intraperitone-
ally for rats. Guest and Rawson (1939) recommend nembutal
mtraperitoneally for rats. Nutter (lOl-l) considers that too low
niuscle glycogen values are obtained if the animal is decapitated,
as glycogen is then consumed in the muscle spasms. She recom-
mends, instead, that amytal should be given intravenously and
states that in this way she finds less deviation in the values.

The question regarding the mode of death is closely connected
With that of the effect of narcosis on the glycogen stores, which
IS also of the greatest importance for all experiments of long dura-
tion under narcosis. It will, of course, always be necessary to show
that the changes observed in these experiments are not due to
the

narcosis.

Croftan (1908) states that during an ether narcosis of 10 — 15
minutes he can perform laparotomy without affecting the liver
glycogen. Evans, Tsai and Young (1931) likewise state that a
short ether narcosis for cats after a fast of 16 — iS hours does
not affect the liver glycogen. Macleod and Pearce (1911), on
the other hand, consider that the irregularities in the deposition
of glycogen in the different parts of the liver are accentuated
under ether narcosis.

It is a generally known fact that the blood sugar rises under
ether narcosis (See, for example, Epstein, Reis and Branower,
1916, Epstein and Aschner, 1916). Chrometzka and Beut-
MAN (1940) state; »Selbst bei vorsichtiger Atherapplikation kann
es bei entsprechender Narkosedauer zu Blutzuckersteigerungen
hochstens Ausmasses kommen. Bei kurzdaueruder Athernarkose
zum Zweck der Narkosevertiefung ist der Blutzuckereffekt ge-
ringer und fliichtiger®. Most authors also find that ether nar-
cosis tends to reduce the liver glycogen. Laubbr (1938) states
that the decrease in liver glycogen is approximately proportional
to the duration of the narcosis, whilst Laubbr and Bersin
(1939), as the effect of an hour’s narcosis on a rabbit, found
». . . dass der Glykogengehalt der Leber wahrend der Narkose um
etwa 50 % sinkt>. Daoud and Gohar (1933 — 1934) consider
that ether narcosis reduces the liver glycogen by 40 — 50 %.
Evans, Tsai and Young (1931) consider that amytal narcosis
after a few hours markedly lowers the content of liver glycogen.
About 50 % are lost in two hours. Hines, Leese and Barer
(1928), with a differently planned investigation, had previously
arrived at the same result. Edeund (1942) states; ^Eine massig
Starke einmalige Dosis Narkotal verursacht keine Veranderung
des Glykogen — und Fettgehaltes der Leber*.

Steenmetzer and Swoboda (1928), summing up, state; *Alle
untersuchten Narkotika bewirkten eine Hyperglykamie . . . Die
dutch die Narkotica erzeugte Hyperglykamie wird als eine Ent-
hemmungserscheinung auf das im Hirnstamm gelegenene Zucker-
zentrum aufgefasst.* Eiseer and Hemprich (1932) report that
3>bei Luminalgaben in nicht todlichen Mengen wurde stets eine
leichte Erniedrigung beobachtet*, namely in the blood sugar.
Hrubetz and Beackberg (1938), after testing a number of

barbiturics, arrived at the result that »with each of the barbiturics
studied there was a marked depression in the glycogenolytic
power of the liver . . .»

CORI (1931) writes: »...one must be aware of the fact that
all anaesthetics so far available, including amytal, have a de-
pressive influence on the glycogen formation in the livers- .

Evans, Tsai and Young (1931) discuss what bearing the loss
of body heat during narcosis may have on the decrease of the
liver glycogen, and Tachi and Takai (1926) consider it to be of
fundamental importance.

Reports regarding the effect of narcosis on the muscle glycogen
are found but sparsely in the literature. Moscati (1907) categori-
cally states ». . . jedenfalls kann der Chloroform, das die Muskeln
zur Erschlaffung bringt, im Gegensatz zu manchen anderen che-
mischen Stoffen bei nicht zu langer Dauer der Einwirkung kein
Einfluss auf den Glykogengehalt zugeschricben werden.» Schenk
(1923), having studied the effect of chloroform narcosis on dogs,
states: »Der Glykogenvorrat des Muskels ist am Ende der Nar-
kose und in den folgenden Tagen herabgesetzt>. He supposes
that chloroform causes cellular damage which tends to retard the
glycogen synthesis. Daoud and Gohar (1933 — 1934) found the
muscle glycogen unchanged after ether narcosis, whereas the liver
glycogen was reduced by 40 — 50 %. HiNES, Leese and Barer
(1928), having studied the effect of glucose infusion on dogs under
amytal narcosis, state: sit was found that approximately the
same increase in muscle glycogen had occurred in animals with
and without amytal anesthesias.

It thus appears from the literature that all narcoses of consider-
able duration result in a reduction of the liver glycogen. Reports
regarding the effect of a short narcosis are more sparse, but the
general view seems to be that their effect is less or none. The
reports regarding the reaction of the muscle glycogen are not
very enlightening.

In my own investigations I adopted only a very brief narcosis.
For my analyses of the . initial postmortal glycogenolysis (see
below), I found it necessary to anesthetize the animals, so that
they could be immediately dissected. Thus, the effect of the nar-
cotic lasted merely a few minutes. In the table Fig. 14 mice which

Fig. H. Table. The effect of avertin anesthesia on the glycogen content
of the body. Males. Food ad lib. Animals killed at 3 -p. m. The animals
were given 5 mgm. avertin by intraperitoneal injection. The anesthesia
was induced in SO — 60 secs., when the animals were immediately decapit-
ated and dissected in the usual way. The control animals were decapitated
unanesthetized and dissected in the same way.

Controls

X±£-

Avertin

animals

x± e-

d±£^

Liver glycogen
per cent

1.84 ±0.29
(n=ll)

1.68 ±0.32
(n = 11)

0.16 ±0.43

0.372

0.8— 0,7

Body glycogen
per cent

o.n8±o.oi4

(n=ll)

0.114±0.020

(n=ll)

0.004±0.024

0.167

0.9— 0.8

Mgm. glycogen
per JO gm.
weight minus
alim. canal and
spleen.

21.9 + 2.8
(n = 11)

20.4 ± 3.8
(n = 11)

1.5 ± 4.7

0.319

have been decapitated and dissected in the usual way are com-
pared with mice which have been decapitated and dissected under
amytal narcosis. They first received an injection of avertin in-
traperitoneallj' and, when after about 30 seconds to 1 minute it
had induced narcosis, the animal was decapitated. As shown by
the said table, there is no difference whatsoever between these
two groups in regard to liver or muscle glycogen.

Many investigators have in fact adopted this proceduce without
preliminary testing and have taken their samples of organs under
a brief anesthesia of some kind. On the other hand, especially in
the case of large experimental series, this procedure entails no
advantages as compared with simple decapitation.

II. Poslmortal glycogenolysis is a phenomenon of funda-
mental importance for an investigation such as this. It is, of
course, essential to make the analyses as soon after death as
possible, so that the values are not changed by postmortal gly-
cogenolysis. Whether this is technically practicable or not de-
pends on (1) how soon the postmortal glycogenolysis sets in and
(2) how soon the samples can be prepared. From earlier authors wc

find slalcments to the effect that it had taken them about 30
minutes to one hour to prepare the organ in question before lay-
ing it in hot KOH. In accordance with my own technique, the
sample for liver analysis lies in hot KOH within 50 seconds, and
the muscle sample within 90 seconds, after decapitation. Greater
rapiditj' can scarcely be attained in practice.

The second above-mentioned question as to how soon the
postmortal glycogenolysis sets in now remains to be discussed.

A. The postmortal glycogcuolysis in the liver. The reports on
this subject in the literature show great divergence. Claude
Bernard (1855), in his earliest experiments, ascertained that the
liver develops sugar postmortally and in course of time he dis-
covered that this sugar is formed by the decomposition of gly-
cogen. Many earlier authors, such as Boehm and Hoffman
(1880), Kulz (1881), and Garnier and Lambert (1897), state
that they had observed a slow glycogenolysis in the liver.
Lubarsch (1906) mentions »eine ziemlich rasche Zersetzung des
Glykogens sobald das Leben der Zellen unterbroehen isti>.
Meixner (1911) found that within one hour post mortem
16 — 54 % of the liver glycogen disappears in rabbits and 100 %
in guinea-pigs. IMacleod and Pearce (1911) showed that the
postmortal glycogenolysis in the liver sets in within the first
20 minutes after death, and that it has its greatest intensity in
the eourse of the first hour. Afterwards, they state, it continues
at a fairly constant rate for several hours. Evans, Tsai and
Young (1931) state that no glycogen is lost during the first 40
seconds. Setting out from this value, they find a gradual diminu-
tion, so that after 8 minutes the liver contains only 60 per cent,
of the initial glycogen value. Bobbit and Deuel (1940) find
that the glycogenolysis in liver substance kept outside the body
at 37° C is »much slower than- generally believed*. They state
that the intensity of the glycogenolysis varies in different animals,
being most marked in rats and diminishing in the following order:
pinea-pig, rabbit, dog. On the basis of their relatively small
investigations, these authors report that the liver glycogen in one
hour diminished by 31 % in rats and 13 % in rabbits. Bomskow
and V. Kaulla (1942) emphasize the importance of rapid dissee-

Fiff. 15. Table. The ■postmortem glycogenolysis. The glycogen content of
liver pieces taken from the same liver lobe of normal rabbits at different
times after decapitation. Each value is given as percentage of the glycogen
content of the first sample, which lay in hot KOH within 40 secs, of de-
capitation. The livers were then placed in a damp
chamber at room temperature.

Time after decapitation

No. of animals

Hemaining glycogen
per cent of first sample

x± e-

— ‘'X

40 secs.

100 + 0

5 min.

94,+ 4

15 »

84 + 5

60 »

71 + 5

150 »

64 + 5

240 .

60 + 5

tion in order to avoid the complicating effect of the postmortal
glycogenolysis. They consider it to be sufficient if the sample
of the tissue is placed in the hot potash lye within minutes.

In short, the reports in the literature regarding the postmortal
glycogenolysis in the liver are by no means in good agreement so
far as relates to its intensity and the time at which it sets in.

My own investigations were made on rabbits and rats. The
first sample of liver was taken immediately after decapitation,
and the piece lay in hot KOH within 40 seconds. The rabbit liver

Fig. 16. Table. The post-mortem glycogenolysis. The glycogen content of
liver pieces of normal rats at different times after decapitation. Each
value given as percentage of the glycogen content of the first sample,
which lay in hot KOH within 40 secs, of decapitation. The livers
were then left in situ at room temperature.

Time after decapitation

No. of animals

Hemaining glycogen
in per cent of first sample

X + £5

40 secs.

100 ±0

5 min.

86 ±8

15 >

62 ±4

75 >

42±3

Fig. 17. Diagram showing the ’post-mortcm glycogenolysis in the liver.

VemainJn^ tn p^r csnt

cf gf if cogsn cent tot ef f fret sarnpfc

0 53 t33 !53 230 250 fhns

in the rabbit,
in the rat.

Fig. 18. Diagram showing the post-mortem glycogenolysis in the liver.
Lines of regression constructed according to the
method of least squares.

Pefn*in/f>^ jtyco^mn !n per cent
cf^ytegen content effirzt umpk

“ rat.

was then dissected and placed in a moist chamber at room tem-
perature. New samples were taken at intervals (see Fig. 15). The
studies on rats were made in the same manner, except that the
liver, after the first sample, was left in situ at room temperature.
In order to prevent evaporation, the abdomen was closed in the
interval between the samplings. The results are tabulated in

Fig. 16.

The course of the glycogenolysis can be read from the diagrams
in Figs. 17 and 18. It will be seen from Fig. 18, where the line of
regressions has been constructed according to the method of least
squares, that the postmortal glycogenolysis in rat liver starts
1.8 minutes, and in rabbit liver 2.6 minutes, after deeapitation.
This, of course, is correct only on the assumption that the
glycogen content in the first sample is really representative of
the liver glycogen at the moment of decapitation, i. e. that no
glycogenolysis has occurred during the 40 seconds it takes to
prepare the first sample. This question will be discussed under
the section on the initial postmortal glycogenolysis (p. 53).

The glycogenolysis sets in earlier and has greater intensity in
the rat than in the rabbit. Bobbit and Deuel’s unproved state-
ment has thus been verified. The diagrams on Figs. 17 and 18
also show that the postmortal glycogenolysis, as already pointed
out by Macleod and Pearce, is most rapid in the course of the
first hour, indeed, during the first half-hour after death.

In order to verify the figures thus found for the time at which
the glycogenolysis sets in, tests were also made with double liver
samples as soon as possible after decapitation. The first sample
lay in hot KOH within 40 — 50 seconds, the second within 30 — 40
additional seconds, after decapitation. Both of them had thus
been placed in hot KOH before the time at which the glycogen-
olysis, according to the above data, should set in. The result of
these analyses is tabulated in Fig. 19. According to these investiga-

Fiff. 19. Table. Glycogen content of two sticcessive liver samples, taken as
soon as possible after decapitation. The first one was laid in hot KOH
within JfO — 50 secs, of decapitation, the second SO — 40 secs, later.

Animal species

No. of
animals

Glycogen content
per cent

d ± £5

Rat

2.18

2.21

0.08 ± O.OS

Rabbit ser. I

7.84

7,77

— O.07 +. 0.07

> > II

7.16

6.99

-0.16 ±0.16

. » in

2.18

2.21

0.03 ± 0.03

tions, there is no systematic difference between the glycogen
content in the first and second liver sample, that is to say, no
loss of glycogen had occurred in the interval between the two
tests.

B. The 'postmoTixil glycogenolysis in muscles. This question has
received less attention than has been given to the same proeess
in the liver.

Boehm (1880), unlike a number of earlier authors, found no
diminution of the musele glycogen within 2 hours. Kulz (1881)
states that the proeess is very slow. Cramer (1888) finds a
substantial deerease in the eourse of 4 hours and stresses the
importanee of rapid dissection. Moscati (1907), working with
operative material from man, eonsiders that »der Anfangswert
fiir den Glykogen sinkt nur langsam ab, eine Stunde nach dem
Absterben verandern sieh nur die dritte oder zweite Dezimale*.
McKay (1928) considers that glycogenoly.sis in intact muscle
proceeds slowly. CoRi (1931) writes: sin the first few seeonds
after stunning the animal, glycogenolysis is very rapid, but as
the lactic acid content of the muscle rises to higher and higher
values, glycogenolysis is slowed down more and more and eventu-
ally comes to a standstill. Since the sampling of the muscle
generally falls into the latter period, the error introduced by the
loss of glycogen is apt to be relatively constant in the different
experiments. For some investigations it may therefore not be
absolutely necessary to determine true glycogen values, but it
would seem preferable to avoid the possibility of error and hence
of wrong conclusions.* CoRl (1930) recommends that, after vivi-
section under amytal narcosis, the muscle should at once be laid
in hot KOH.

This view as to the very rapid initial postmortal glycogeno-
lysis is based mainly on the studies of Davenport and Daven-
port (1928), who show that the only way of obtaining low lactic
acid values in muscles is immediate freezing of the whole intact
muscle. Even if the freezing is not deferred for more than 2 — 6
minutes after death, these authors find a substantial increase in
the content of lactic acid, which is considered by them to be
produced by the decomposition of glycogen. These authors attach

Fig. 20. Table. M. triceps sin. of rabbit was dissected immediately after
decapitation. It was divided into two portions, which in rapid succession
were weighed and placed in hot KOH. The first one lay in hot KOH
within 50 secs, of decapitation, the second within
additional 20 — 30 secs.

Sample I

Sample 11

X + e-

d±£d

Glycogen content per cent

0.39

O.tl

0.02 ± 0.03

very great importance to muscle spasms and the manipulations
during dissection as causes of the rapid decomposition of the
glycogen, Goldblaott (1933), on the other hand, considers that
uniform and reliable values for the muscle glycogen can be ob-
tained on analysis for up to one hour after death, Ajuderson
and Macleod (1930) state that >no change occurs in the amount
of glycogen in intact ^mammalian muscle as a result of standing
at room temperature for 1 hour after death*, Simpson and
Macleod (1927), on the other hand, arrive at the result that
»in mammalian muscle which is frozen and ground up in liquid
air immediately after its removal from a living animal (decapitated
cat) and then allowed to stand at room temperature, glycogen
almost entirely disappears within 20 — 30 minutes*. The disparity
in the results may be explained by the fact that glycogenolysis
in injured tissue is much rapider than in intact tissue (Evans,
Tsai and Young, 1931, and others).

Thus, the reports in the literature are divergent also in regard
to the postmortal glycogenolysis in the muscles. As this question
is of the greatest importance for the present study, a control
investigation of it seems necessary.

My oion investigations comprise two series. In view of CORl’s
above reported views, based on the studies of Davenport and
Davenport, a series of investigations were first made on musculus
triceps brach. sin. of normal rabbits. The muscle was dissected
as soon as possible after decapitation, and was divided into two
pieces, which were weighed and laid in hot KOH, the first piece

Fig. 21. M. triceps sin. of rabbit was dissected immediately after decapit-
ation and divided in two. The one was immediately weighed and placed
in hot KOH within 50 secs, of decapitation, the other kept for 15 mins
in a damp chamber at room temperature before
being placed in hot KOH.

Muscle glycogen content

Time after decapitation

No. of animals

per cent

X ±£-

50 secs

0.28 ±0.05

15 mins

0.28 ±0.01

within 40 — 50 seconds after decapitation and the second
within 30 — 40 additional seconds. The results of these tests are
tabulated in Fig. 20. We see from this table that the mean differ-
ence between the glycogen content in the two pieces is less than
its standard error. The correspondence in regard to the amount
of glycogen is thus good and no systematic deviation is found.

In the second series the muscle was dissected in the same way
and divided into two pieces. One of them was immediately weighed
and laid in hot KOH within 50 seconds after the decapitation.
The other piece was kept in a moist chamber at room tempera-
ture for 15 minutes before it was weighed and laid in the hot
KOH. The results of the analyses are tabulated in Fig. 21, which
shows that the amount of glycogen is the same in both pieces.
Thus, no glycogenolysis had occurred during the interval between
the first test and the second. According to CoRl, however, the
glycogenolysis is most rapid during the first seconds after death,
after which equilibrium is attained. According to the reports of
Davenport and Davenport, longer intervals of time must have
been involved than those in question here (see the immediately
following section).

C. The initial postmortal glycogenolysis. The entire above
reasoning is valid only under the assumption that the amount, of
glycogen in the sample plunged into hot KOH within 50 seconds
after decapitation is really the same as in the organ at the moment

Fig. 22. Table. Normal mice. Females. Food ad lib. The mice were
alternately decapitated and dissected in the usual way and Mlled under
avertin anesthesia, the alimentary canal removed and the remainder cut
into -pieces, which were dropped directly into hot KOH.

Total glycogen
expressed in
per cent of
body-weight
minus alim en-
lary canal
and spleen.

Controls

X + £-

Avertin

animals

X± £-

d + a.

0.184 ±0.027
(n-9)

0.18G± 0.014
(n=9)

0.002 ±0.031

0.066

of decapitation. In order to throw light on this question, a number
of additional experiments Avere made.

These experiments were made on mice narcotized with avertin.
Thej’^ received an injection of 5 mg of avertin intraperitoneally,
on which dose they lose consciousness in about 30 seconds to
1 minute. According to my previously reported investigations on
the effect of short avertin narcosis on the glycogen stores of the
body (see Fig. 14 p. 46) an avertin narcosb of this duration has
no effect AA'hatever on the glycogen contents in the organs.

It has been presumed that the glycogen left in the organs after
a fast of 24 hours is retained with great tenacity, whereas the
glycogen in non-fasting animals is very easily attacked and
rapidly disintegrated. For this reason, the initial glycogenolysis
was studied on animals Avith and Avithout 24 hours fasting.

1. On a series of normal mice under avertin narcosis the ali-
mentary canal Avas quickly cut out, Avhereupon the remainder of
the animal was cut into four pieces, Avhich Avere directly plunged
into hot KOH. This animal series Avas compared with concurrent
series of animals Avhich were decapitated and dissected in the
ordinary Avay without narcosis. The results are tabulated in Fig.
22. The glycogen values show very good correspondence. In
these tests Avithout fasting, the liver contains large amounts of
glycogen.

Fig. 23. Table. Normal mice. Males. 20 hours fasting. Killed at 8 a. m. The
controls were decapitated and dissected in the usual manner. The animals
of the test series were anesthetized with avertin and cut alive into pieces
which were dropped directly into hot KOH. Earlier experiments showed
that the alimentary canal after this period of fasting does not contain any
polysaccharides which might affect the result.

Controls

Avertin animals

X + e-

Glycogen content in per cent .

0.042 + O.003

0.042 + 0.002

(n = 11)

2. After 24 hours fasting, the liver in mice contains merely
insignificant amounts of glycogen. A glycogen analysis made in
the same way as in the preceding series will thus be in the first
place representative of the rest of the body, notably of the
muscles. In preliminary investigations it was moreover shown
that the intestinal lumen in mice after 20 hours fasting does
not contain any polysaccharides, whence also the intestine under
these conditions may be included in the analysis. In this series
the entire avertin-narcotized animal was cut into four pieces,
which were directly plunged into hot KOH. This series was
compared with the concurrently studied series of animals, where
they were decapitated and dissected in the usual manner without
narcosis. The results are tabulated in Fig. 23. The table shows
that there is no difference in the glycogen content between the
two series. As a supplementation of this series, an experiment -was
made in which the animals were decapitated after 24 hours fasting
and then immediately cut into pieces, which were plunged into hot

in which they lay within 12 seconds after decapitation.
Also these animals were compared with a series of animals which
had been decapitated and dissected in the usual way without
narcosis. The results, which are tabulated in Fig. 24, show good
correspondence between the glycogen values of the tw'o series.

3. In different quarters stress has been laid on the importance
of rapidly interrupting the post-mortal glycogenolysis, and
attempts have been made to attain this in various ways. For
example, the dissected organs have been immediately frozen with

Fig. 24. Table. The initial ■post-mortem glycogenolysis. Normal mice.
Males. 24 hrs. fasting. Killed at 5 p. m. The animals of one series immedi-
ately after decapitaiion were cut into pieces and dropped into hot KOH
within 12 secs. The controls dissected in the -usual manner.

Glycogen con-
tent in per
cent

Controls

x±e-

Animals cut
to pieces
within 12 secs

X±£-

d±£^

0.041 ±0.004
(n - 10)

0.O44 ±0.007
(n = 10)

O.003±O.008

0.376

■1

■B

carbonic acid snow or in liquid air, (Sahytjn and Luck, 1929),
which, however, did not show other glycogen values than if the
pieces had been laid direct in hot KOH.

The most rapid way of interrupting the glycogenolysis in small
animals seems to be to kill them by casting them into liquid air.
A mouse in that case will be frozen in 10 to 15 seconds. In my
experiments I used mice which after 27 hours fasting, were cast
into liquid air. When they had frozen they were cut in pieces
and laid in hot KOH. In that way all processes in the organs
had been interrupted as rapidly and effectively as possible. These
animals were compared with the control series in which the ani-
mals, after fasting for 27 hours, had been decapitated and immedi-
ately after decapitation had been cut into four pieces, which were
plunged into hot KOH, As the liver in mice after this fasting
period contained merely a very minute amount of glycogen, the
analysis results obtained are practically representative of muscle
glycogen. The are tabulated in Fig. 25. This table shows that
there is no marked difference between the two animal series.
The probability that a difference exists is P = O. 05 . This result
seems to indicate the desirability of further investigations. It
should be noted, however, that it is the animals killed by being
thrown into liquid air that show a lower amount of glycogen than
those treated in the usual way. Consequently, these tests have
already shown that freezing with liquid air at any rate does not
involve any advantage as compared with decapitation and dissec-
tion in the usual way.

Fig. 25. Table. Comparison between animals Jcilled by being thrown into
liquid air and animals Jcilled by decapitation, cut into pieces and dropped
into hot KOH, The jrozen animals were sectioned and placed
in hot KOH. Mice. 27 hrs. fasting.

Treatment

No. of
animals

IVeight

Glycogen content in
per cent

Liquid air

13.5

0.071 ±0.003

Decapitation

13.6

O.079 ±0.002

d = 0.008±0,004
t = 2.000
df = 30

P = 0.05

A matter which must be taken into account here is the question
whether the hot potash lye penetrates into a tissue as rapidly
as it thaws, as otherwise there is every prospect of obtaining in
the interval a rapid glycogenolysis. It has in fact been shown,
for example, by Simpson and Macleod (1927) and Evans, Tsai
and Young (1931) that the rate of the glycogenolysis is greatly
increased in damaged tissue.

Summary.

These investigations show that no difference in the amount
of glycogen can be observed in mice treated as rapidly as is
possible and in mice treated in the standard way adopted through-
out in this investigation. As the fasting animals in the above
reported investigations have a very low content of glycogen in
the liver, the glycogen values are practically representative of
the muscles, and consequently there is no change in the glycogen
content of the muscles during the first 90 seconds after decapita-
tion. Aecording to previously reported studies of glycogenolysis
in rabbit muscles, there was no difference in glycogen content in
the organs analyzed 50 seconds after decapitation and those
which had not been analyzed until 15 minutes later. CoRl’s view
regarding an extremely rapid initial post-mortal glycogenolysis
has .thus not been verified.

SYUO-ft

;t\ra®

jyyIs

tViAis

OAlt

•vio^^

ihe

atv^T^’

.a^s

ond reftati^^’

deca?'^^" eg«d. uea'-'" *e aa®', *e

*'sa®* ®„olsa'a “‘'^'''til^aai 8'S“^s \o ^■- ^ ,,as

ot «> V^tt cva'''

-ost-®", .„5i as » ^ ^„v„ecl ^,ea ® _ aBaes®'“^^^pda-

J>te *“!: *e

a®®" °'

^ V,oa®. >" : w '='> "“ ‘ tes "O'- “ .mMo" °' tiet 4=®?’

l\oa a'-

CHAPTER 7.

The hearing of the animals’ age an<l sex on
the ainonnt of the glycogen depots

A. Sex. In tlie earlier liLeratnrc \vc find no reports on the
bearing of sex on the amount of the glycogen depots. On the other
hand, since the middle of the nineteen-thirties this problem has
been attacked in regard to the liver by a number of investigators.

GniJisilKIMEn and Johnson (1030) could not show any differ-
ence between the sexes with respect to the amount of liver
glycogen in rats. Deuee, Gueick, GntiNEWAED and Cuteee
(lOSi) arrived at the same result as regards rats and guineapigs,
whereas Gulick, Samuees and Deuee (1934') found somewhat
lower liver-glj'cogen values in females than in males when the
animals after 48 hours fasting received 5 mg. of glucose per em-
body surface and were afterwards fasted for another 48 hours
before the analysis. BEATiiEmviCK, BnoDSUAW, CuEEmoKE,
Ewing and Larson (193G) state, in regard to the glycogen con-
tent in the liver and muscles in their normal material of rats:
»The sex difference in the glycogen content of their tissues was
clearly apparent». In regard to the liver, the males had undoubt-
edly higher glycogen values. As regards the muscles, however, the
difference was less convincing. Deuee, Hallman and Murray
(1937) show higher liver glycogen in male than in female rats and
Deuel, Butts, Hallman, Murray and Blunden (1937) state
that »the level of liver glycogen in the females averaged only
GO per cent of that in the liver of the males*. In the young ani-
mals, however, they found no sex differences. Deuel, Hallman,
Murray and Samuels (1937) consider that the deposition of
glycogen in the liver of rats, on administration of glucose per
os after fasting, is larger in males than in females. This, they state.

is not due to the difference in resorption, as the females under
these conditions resorb much more rapidly than the males. The
difference is considered by them to be due to greater oxidation in
the females. Merten (1939), ^Die weiblichen Tiere (rat) wiesen
stets einen niedrigeren Glykogengehalt der Leber aufi'. Grayman
( 1941), confirms the view of Detjee et alios.

As regards the rabbit, Sjogren, Nordenskjold, Holmgren
and Mollerstrom (1938) have found that the males have a
higher liver glycogen content than the females. Moreover, the
females lose their glycogen more rapidly than the males during
fasting.

Neufeld and Collip (1941) could not show any sex differ-
ence in the liver glycogen content in mice.

A matter of the greatest importance for investigations of this
nature is the variation in the series. Daoud and Gohar (1933)
found that male rats were more sensitive than the females to
factors which might affect the contentof liver glycogen. NeufelI),
ScoGGAN and Stewart (1940) as well as Neufeld and Collip
( 1941) state that they have obtained more regular figures for
females than for males, but that the average value is the same.

Ovm investigations. For my studies on rabbits it was im-
possible to arrange sufficiently large series Avith animals of one
sex only, I have therefore tried, instead, to have an equal number
of males and females in each experimental series, in order to
obtain comparable values even if there should be any sex differ-
ence. From the animal material in the normal series males and
females Avere picked out just for this special question and Avere
compared AAuth one another. The results are tabulated in Figs.
26 and 27. These tables show no difference in the amounts of the
glycogen depots in the males and females.

For mice and rats experimental series were arranged for direct
determination of possible sex differences. The results are tabulated
in Figs. 28 to 29. As Avill be seen from these tables no sex differ-
ence could be shoAvn in regard to the liver and muscle glycogen
in my material of non-fasting rabbits, rats and mice. On the
other hand, the standard deviation in the series is somewhat less for
females than for males, just as Daoud and Gohar, Neufeld,
ScoGGAN and Steavart as well as Neufeld and Collip had

Fig. 20. Table shmcing the storage of glycogen in male and female rabbits and rats. Normal rabbits, 2—0 hrs. fasting,
deeapitated at 0 — S p. m. Normal rats, 2 hrs. fasting, decapitated at U a. m-.

r3 «

<!r-

o o
O O

o o

• d d

“ u +i

+i +i

+I4i

■p B. [>'•

: « «

o d

i d d

o ei

oe ac-

8 c ’x
^

“ ^ +1

d d
4-| 4|

d d

-H

^ cc

^ !X

•r «

ci ci

!>• «0

« Cl

00 ^

'u to

d d

CO •?*

^ I. +!

-H -H

-H +!

^ ix

o o

M O

c: *4

•*< d

to to

«->

CO o

J n

" l+i

■H -H

-H -H

»A

^ CO

f l«

(M — «

-fH t-N

•?♦ C3

to lO

tH ^

o "3

d .§

Cl Cl

^ c

o "b

c/3

Animal

species

;

t£

•xj -S’
£ fa-

C3 ^

••• •>»

:§ ^
2

e «

5 ®
g

c?>

*?*

t«

“

^ .

to"

-H

t-*.

•3

>»

»p«

»a

•o

■H

-H

■H

-tJ

.*3

(3 C3

« «

already pointed out. I have therefore used only females in the
principal series for mice.

In the literature it will be found that most authors, if they
had reflected at all about this question, had used only males,
without further investigation of the conditions. The reason is
stated to be that in this way they believe that they escape from
the possible effect of the female sexual cycle. BokeLmann and
Dieckmann (1930) as well as Bokemlann, Dieckmann, Ka.uf-
MANN and ScHERlNGER (1931) state that they could show a
connection between the glycogen reserves of the animal and
sexual cycle, but their differences are not of such a magnitude
as to be statistically significant. No connection between these
factors has hitherto been demonstrated.

B. Age. Since Claude Bernard (1859) examined fetus livers
and was unable to show any liver glycogen until the second half
of the embryonal life, a number of investigations of fetal liver
and muscle glycogen have been made. The earlier studies were
made with more or less defective technique and therefore had
scarcely any decisive importance for the question when glycogen
first begins to be deposited and in what quantities it then occurs.
Barpurth (1885) states that, in examinations of fetuses, he had
not observed any glycogen in the liver, but had found it in several
other organs of rabbit, sheep and guineapig in different stages of
development. In homo fetuses which had died during delivery
Salomon (1874) found an abundance of glycogen, whereas
Marchand (1885) was unable to observe any. Nor was the latter
able to find any glycogen in a lamb aged a few months, 15 minutes
after death.

Pfluger (1903), who Avorked with a better technique, observed
liver glycogen during the first half of the development of the
fetus, likcAvise Sundberg (1924) with a histological technique.
Similar reports are found in Adamoff (1905). Livini (1927)
detected traces of glycogen in the liver of a homo fetus 18 mm
in length.

Even if there is some divergence of opinion as to when the
liver first begins to deposit glycogen, it is generally agreed that
the glycogen content of the liver is at first low, even lower than

that of the muscles. (Pfluger 1903, Gierke 1905,
and Cramer 1906, Livini 1927, Stieve and Kjipsv 19371. and
others.) The amount of liver glycogen is considered , to fee ' low :
at first, but increases considerably towards the end of fetal life
(Lockhead and Cramer, Livini, Stieve and KAPs).vr ; '

The muscle glycogen content, on the other hand, is high: in the
fetus already at an early stage (Gierke, Ppleger, Sendberg); .

Thus, to judge by the literature, the liver seems at first to con-^'
tain merely small amounts of glycogen, at any Tate not more than ;
e, g. the skeletal muscles. It is not until just before birth that
the glycogen content rises. In other words it is hot until' this
stage that the liver begins to act as a store for glycogen. : • : ; .

According to Stieve and Kaps, the glycogen content of the
liver during the first weeks of extra-uterine life lies at about
the same level as at birth. Hax (1927), who worked with puppies
considered that they had a liver rather poor .in glycogen at birth
and that the glycogen content increased during .the suckling
period. Jonen (1924), who studied new-born dogs, ’found' an
average glycogen content of 5 % in the liver and. 0.6 % in the
muscles. y.',;:-..

Deuel, Butts, Hallman, Murray and Blunden (1937);,
studied the liver glycogen in rats of different ages. All the animals^
were put on the same diet and were killed at 8 — ^10 a.' hi.' during
a short period in the spring. They state: ^The level of glycogen
present in the liver of unfasted rats is highest 'at 39—40 days,
at which time it exceeds 8 per cent. From this level' it ^adually
drops to a value of about 4 per cent, which is found in male rats
75 days of age. Approximately the same values were 'noted in rats
19 — ^24 months old. There was no sexual difference noted in, the
liver glycogen of rats 26 — ^29 days of age nor in old rats (17 to 24
months). In the other groups the level of liver glycogen in :,the
females is constantly lower than that of, the, males. The liver
glycogen reached a minimum value in the females which , were
three months old. » ■ .. . y ■

Heyman and Modic (1939) studied rats aged 8.^11 rdays;
10 — 12 weeks and 2 — 3 years. All the animals were: killed at' the'
same time of the day after a fast of 48' hours. >The liver glycogen
values for rats 8 to 11 days old is dne-third That for rats in the

two older groups. During the first 6 weeks of post natal life this
value increases steadilj% the average values for the animals 6 weeks
old being 5.5 gm per 100 gm of fresh liver, which is only slightly
higher than the average value for the full-grown animals . . . The
increase in content of glycogen in the liver of baby rats takes
place mainly during the nursing period . . . The muscle glycogen
values in baby rats are certainly not lower than those in the two
older groups, if they differ at all . . . Fasting for 1 and 2 days
diminishes the content of glycogen in liver and muscle of rats
in all three groups with the same ease. This does not support
the hypothesis that the liver retains its glycogen depots with
greater tenacity during infancy than during later periods of life.»
Bomskow and v. KIaulla (194'2), after experiments on guineapigs,
likewise stress the bearing of age on the amount of the glycogen
depots. Those authors also find a rising content of muscle
glycogen with advancing age. The series in their work are, how-
ever, too small for their figures to be quite convincing, in view
of our knowledge of the considerable individual variations.

To judge by the literature, the investigators seem more or less
consciously to have evaded the source of error involved in possible
differences in the glycogen content with varying age. Without any
discussion and without giving reasons, they have taken animals
of approximately the same size, i. e. age, which indeed seems
quite natural. In correspondence with this, Guest and Rawson
(1939) recommend very narrow age limits for animals used for
studies of the glycogen depots. If the animals’ weight is adopted
as an indicator of their age, however, regard must be paid to the
fact that e. g. female rats of fertile age grow more slowly than
males (Donaedson, 1908).

Oum experiments. The liver and muscle glycogen values for
rabbits, mice and rats of different ages were compared. The rabbit
senes for these age studies were composed from different control
series, it having been found, as explained in the preceding section,
that sex does not play any part in the glycogen content of the
rabbit. As regards mice and rats special series were arranged for
this purpose. The animals in the different age groups were treated
m exactly the same way and were studied at the same time.
The procedure was the same as in the principal series. The results

Fig. 30. Table showing the storage of glycogen in rats and rabbits at different ages.

Fiff. JJ. Tabic ,^hou-ing the storage nf r;l;irnprn in mice at tlSJfcrnit n^a.

0 o o

1 i !

o c o

£.E “

H
c- ^

K.sr* -

4 /

O . 4 J

I ^

—

t« I ■> t •

»f: ei »~

M -H -n

o o t«
ri ci cri

no cc CD

Ci Ci —

c o c

cod

*^1 -M -il

*!% O

e> --• o
o o o

. . o

e o c5

I- c» r:

o' d ci

are tabulated in Figs. 30 — 33. As will be seen from these tables,
the two rat series show a very good confirmity in the content
both of liver glycogen and muscle glycogen. As regards the rab-
bits, on the other hand, there is a statistically significant differ-
ence (P == < O.ooi) between younger and older animals, in that the
younger show a considerably lower liver glycogen content than
the older ones. As for the muscle glycogen no such difference
could be observed, and the two series show a good correspondence.
As regards mice, the bearing of age on the glycogen depots was
investigated in three different experiments. In the two first, older
and younger animals show a very good correspondence in regard
to the liver glycogen content, whereas in the third series the
younger animals were found to have a smaller amount of liver
glycogen than the older ones (P = O.os — O.02). As such a differ-
ence occurs only in one group and the statistical probability is
not definite, it seems unlikely that this is a systematic difference.
It seems more probable that some external unknown factor has
come into play. As regards the glycogen content in the body
a very good correspondence was always found between the older
and younger animals, as also for the total amount of glycogen
per weight unit.

It may seem curious that age should have a bearing on the liver
glycogen content in rabbits, but iiot in mice and rats. This may
possibly be due to the fact that the differences in age were not
comparable. To judge by the reports in the literature on the
influence of age on the glycogen depots, even if the data are
somewhat disparate, it may be inferred that the depots are gener-
ally considered to be somewhat larger in older animals than in
very young ones.

Summary.

Some investigators state that in rats, guineapigs and rabbits
they have ascertained that the glycogen reserves of the males
are larger than those of the females. As regards mice such differ-
ences could not be found. In my investigations such sex differ-
ences, under the existing experimental conditions, could not be
observed in rabbits, rats and mire. On the other hand, the

standard deviation is somewhat less for females than for males.
On his account, only females were used for the principal experi-
ments on mice. For purely technical reasons, it was impossible
to procure a sufficient number of animals of one sex for the studies
on rats and rabbits. I tried, instead, to secure the same number
of animals of each sex in the series which were to be compared.

The more extensive investigations reported in the literature
indicate that older animals have a larger amount of liver glycogen
than very young ones. In my experiments I observed such an
age difference in rabbits, but not in mice and rats. All the in-
vestigations reported in the sequel were made, so far as possible,
with animals of the same age. This was at any rate the case
with the serial experimental animals and the corresponding
controls.

(59

CHAPTER 8.

The hearing of the diet and tlie length
of the fasting perioh on the
glycogen depots

I. The effect of the diet.

It is obvious that the composition of the diet must have a
considerable bearing on the magnitude of the glycogen reserves,
and attention has long been directed to this matter. The earlier
experiments were chiefly intended to ascertain how far the
glycogen content could be forced up. Thus Schondorff (1903)
states that in dogs on a meat diet high in carbohydrates the liver
glycogen can be forced up to 18.7 per cent and the muscle glycogen
to 3.7 per cent. Pfluger (1907) put dogs on a lengthy fast and
then gave them a monotonous fat, protein or carbohydrate diet.
He writes; j A ls Ergebnis dieser Untersuchungen darf mit grossler
Wahrscheinlichkeit behauptet werden, dass die Leber bei voll-
kommener Entziehung der Nahrung bis zum Hungertode fort-
fahrt, Glykogen zu bilden. Wird der Leber als Nahrung in iiber-
schiissiger Menge entweder nur Fett oder nur Eiweiss zugefiihrt,
so hort die Glykogenbildung auf oder wird auf ein Minimum
heradgedriickt. Wird aber der Leber als Nahrung in uberschiissi-
ger Menge ausschliesslich Traubenzucker zugefiihrt, so nimmt die
Glykogenbildung in aussergewohnlich starkem Maasse zu, wie ja
langst bewiesen ist.» Junkbrsdorf (1921) confirms Schon-
dorff’s results.

In regard to rabbits Liebig (1940) states that the liver glycogen
is reduced if the animals are put on oats or turnip diet, and that
it is normal if they are given chiefly carrots.

Osborne and Mendel (1924) kept rats on a carbohydratc-frcc
diet for several months. The animals grew well. On analysis of

the whole animal a glycogen content of O.oo per cent was found as
compared with O .12 per cent in animals on normal diet, Gheishbi-
iiER and Johnson (1929 and 1930 a) report a definite increase of the
liver glycogen in rats put on a diet rich in sucrose and a definite
decrease if the diet is rich in fats or casein. Greisheimer and
Johnson (1930 b) write: »Feeding 60 per cent of the total calories
in the form of com starch gave a liver glycogen which did not
differ significantly from that on the balanced diet. With 60 per
cent of the caloric value in the form of sucrose or lard, a signific-
antly higher liver glycogen was found, while 60 per cent of casein
gave a significant decrease. The muscle glycogen content on any
of the test diets did not differ signifieantly from that 6n the
balanced diet.* Sahyun, Simmonds and Working (1934) examin-
ed different diets and write: * Glycogen content of the muscles
of the rat under the conditions of the experiments tends to vary
directly with the percentage of carbohydrate in the diet, being
over 50 per cent greater in the animals on a high carbohydrate
diet than in animals on a high protein diet.* Mirski, Rosen-
baum, Stein and Wertheimer (1938) find a decrease of the
liver glycogen if a large part of the calories in the diet are supplied
in the form of casein or meat. The muscle glycogen on the other
hand, was not affected. Stein, Tuerkischer and Wertheimer
(1939) compare experimental animals on a diet the calories of
which were supplied as to 34 per cent by butter, 34 per cent by
margarine, 26 per cent by casein and 6 per cent by starch, with
normal animals the calory requirements of which are covered as
to 70 per cent by carbohydrates. They find that the controls
have nearly twice as high a liver glycogen as the experimental
animals. Holmgren (1944) studies rats put on a diet of bread
or boiled beef. Both groups receive a full supply of vitamins and
salts. *Die Leberglykogenwerte sind bei Kh-Tieren hoher als bei
Fleischtieren. Die Differenz, I .80 % ± 0.54, ist statistisch sicher.*
As regards guineapigs, Bomskow and v. Kaulla (1942) state:
»Bei der Wechsel von Winterkost (Heu, Riiben und Brot) auf
Sommerkost (Gras und Brot) sowohl cine Erhohung der Leber-
glykogenwerte wie auch einc Erhohung der Herzmuskelglykogen-
werte, vor allem aber eine Erhohung der Skelettmuskelglykogen-
werte.*

As regards mice, Mirski, Rosenbaum, Stein and Wert-
heimer (1938) state that the animals put on a wheat diet have
twice as high a content of liver glycogen as animals kept on a
meat diet.

Kaunitz and Selzer (1937) find a lower liver glycogen con-
tent in rats who had been put on vegetable salt-free diet than
in rats who had been put on a meat diet rich in salt. Crabtree
and Longwell (1936) state that 9 % sodium chloride in the
food increases the liver glycogen in rats, whereas 6 % or defic-
iency of salt does not result in any deviation from the normal.
Galli and Raffo (1939) state that intravenous injections of
common salt tend to increase the fixation of glycogen in the
liver. Kobori (1926) studied the effect on the liver glycogen of
potassium and calcium salts in the food and found that the
glycogen content was highest on a high calcium diet and lowest
on a low potassium diet. »Doch scheint der Einfluss kein erheb-
licher zu sein.» The muscle glycogen was found to be unaffected.

As regards the question of the sensitivity of animals to insulin
on different diets, much work has been devoted to studying the
effect of alkalizing and acidifying diets. Also the effect of such
diets on the glycogen depots has been studied, but the results
are not uniform. Thus, for example, Stein, Tuerkischer and
Wertheimer (1939) state that acidotic diet entails a rise in
the liver glycogen content in the winter, but not in the summer,
whereas alkalosis results in normal values. Goldblatt (1927)
gives rats on normal diet some sodium bicarbonate in the milk,
and then finds that the glycogen almost entirely disappears both
from the liver and from the muscles. Much value can scarcely
be attached to most of the experiments that have been made on
this subject. As a rule, like many other experiments in diet, they
are not arranged in such a way that the animals can be regarded
as ^normal and healthy». It is often remarked en passant, in a
foot note or the like, that a large number of the experimental
animals had died, or that they appeared to be ill. Under such
conditions one must obviously expect great variations in the
glycogen reserves.

The effect of vitamins on the carbohydrate metabolism is a
question which has attracted great interest. Their effect on the

glycogen reserves has also been studied. These investigations, how-
ever, in many cases are badly planned and the experimental
material insufficient, for which reasons the results are often vague
and selfcontradictory.

Vitamin A, supplied in excess to rats on a normal diet, accord-
ing to ABelin (1935) increases the muscle glycogen. He considers
that this rise is entailed by the well-known antagonistic effect
of vitamin A to thyroxin, which tends to reduce the muscle
glycogen. Bauereisen (1938) considers that vitamin A in excess
produces an increased fixation of glycogen in the liver. It is
largest when sugar is concurrently supplied. A lack of vitamin A
entails a reduction of the glycogen reserves.

Vitamin B has been the subject of a whole series of investiga-
tions. Funk (1914) considered that this vitamin was of import-
ance for the carbohydrate metabolism. In Bj-avitaminotic pigeons
BiCKEii and Collazo (1923) found a deficiency of glycogen,
whereas Abderhalden and Wertheimer (1932 and 1934) found
the liver of B^-avitaminotic pigeons richer in glycogen than that
of normal pigeons, and this glycogen rise was considered by them
to be directly proportional to the degree of the avitaminosis. This
richness of glycogen is confirmed, among others, by Tonutti and
Wallrafp -(1938). In rats, Fornarolli and BONI (1940) find
a reduction of the muscle glycogen after a period on Bj-free diet,
concurrently with the appearance of other symptoms. According
to Hermann (1939), the liver glycogen content in rats during
the first stage on B^^-free diet shows a rise, but afterwards returns
to the normal value, and falls below it in the final stage. Tonutti
and Wallraff (1938) state that they have found in rats a reduc-
tion of the liver glycogen content in cases of B^-avitaminosis.
Edlund and Holmgren (1941) came to the same results. The
supply of vitamin B^ in exeess, according to Lajos (1936), pro-
duces an increased tendency towards glycogen fixation. Accord-
ing to Edlund and Holmgren (1941) and Edlund (1942), no
definite change in the liver glycogen content could be found after
the supply of vitamin Bj^ in excess.

Vitamin C: Palladin (1924) sBeim Skorbut verschwindet das
Glygoken aus der Leber und am Ende des Skorbuts war das
Glykogen iiberhaupt nicht nachweisbar» (guineapig). Low liver

glycogen content in scorbutic guineapigs was noted also by
Altenburger (1936) and Shimamura (1938).

Several authors have worked with an increased supply of vita-
min C and have studied the effect on the liver glycogen. Her-
mann (1938) finds a marked increase, Morelli and d’Ambrosio
(1938), Terada (1939), Fichera (1940) and others confirm this,
whereas Tcherkes and Rosenfeld (1941) state that a supply
of vitamin C in excess results in a reduction on the liver glycogen,
and Borell and Holmgren (1945) find no change.

Vitamin- D: PiNCUSSEN (1932) supposes that vitamin D affects
the carbohydrate metabolism.

II. The effect of fasting.

That fasting tends to lower the magnitude of the glycogen
resers’es is a matter on which there has long been general agree-
ment, but investigators are disagreed regarding the amount of
this glycogen decrease.

A) The liver. The earlier investigators as a rule work with
very long fasting periods. Luchsinger (1875) states that the
liver in rabbits is free from glycogen after a fast of 6 — 14 days.
Aldehopp (1889), on the other hand, states that the liver of the
rabbit still contains considerable amounts of glycogen after 6 days
fasting and that of the cat after a fast of 14 days. Pfluger
(1902), in studies on dogs after 28 days fasting, found a liver
glycogen content of 4.8 %. He emphasizes, however, the immense
individual variations. In 1907 he stated »dass die Leber bei voll-
kommener Entziehung der Nahrung bis zum Hungertode fort-
fahrt, Glykogen zu bildeni.. Ishimori (1913) finds a marked
decrease of the liver glycogen after 1 — 5 days fasting. JuNKERS-
DORP (1921) in studies on monkeys still found 3.12 grams of glyco-
gen in the liver after 23 days fasting. Liebig (1940) finds a
substantial decrease in the liver glycogen after a fast of 6 — 7
days.

These long periods of fasting naturally entail profound changes
in the function of the liver cells and therefore are devoid of
interest in this connection. Later investigators worked with

shorter fasting periods as a means of making the glycogen content
in the liver more uniform in serial experiments with the object of
inducing changes in the liver glycogen content by different means.

Fisher and Lackey (1925); j-The liver is the first orpn to
lose its glycogen in starvation in both normal and diabetic dog.
The loss in the amount of glycogen is rapid at first but the
glycogen left after a few days starvation is given up very slowly.j-

After studies on rabbits, Markowitz (1925) states: »The in-
ference is very strong from these experiments that starving rabbits
are never glycogen free*. He found up to 3 % liver glycogen in
starving animals. According to Sjogren, Nordenskjold, Holm-
gren and Mollerstrom (1938) a substantial reduction of the
liver glycogen is found in starving rabbits. The reduction cul-
minates after 48 hours. After 72 hours fasting the liver glycogen
content is higher and the values more irregular. The fast of the
first 48 hours has no effect on the daily variations in the glycogen
content of the liver.

The studies of Agren, Wilander and Jorpes (1931) on mice
show that a 10 hours fast entails a marked reduction of the liver
glycogen in those animals, but without abolishing the daily varia-
tions. Neufeld, Scoggan and Stewart (1940) report a marked
reduction of the liver glycogen after 24 hours fasting and »a fair
constancy ranging from 50 to 70 mg %». After 48 hours fasting
they find less uniformity in the values. Neupeld and Collip
(1941) state in regard to mice that the first 5 hours of the fast
do not entail any reduction of the liver glycogen. Afterwards the
liver glycogen content falls rapidly during the first 14 hours of
the fast, but afterwards keeps approximately constant up to 48
hours. These authors were unable to find any difference in the
sensitivity of males and females in this respect.

A number of studies of this question have been made on rats.
Karczag, Macleod and Orr (1925) found a marked decrease
of the liver glycogen after 24 hours fast. LAWRENCE and McCane
(1931) showed a diminution of the liver glycogen to 0.4 0.3 %
after a fast of 24 — 48 hours. Dittmar (1933) after studies on
rats, states: »Das Glykogen der Leber wird rasch abgebaut, es
sinkt am 1. und 2. Hungertag auf ein Minimum^. Wetzel, Woll-
SCHITT, Ruska and Oestreicher (1936) studied rats on normal

diet. They had merely small series of 4 in each group. The rats
were always decapitated and analyzed from 9 to 11 a. m. These
authors found after

24) hours
2X24- j.
3X24 »

4X24 »

5X24 3.

7X24 »

9X24 s

fast 0.78%

> 0.77 %

1. 0.88 %

2. 0.93 %

» 1.03 %

3 > 1.35 %

* 1.12 %

liver glycogen

Mirski, RosENBAtTM, Stein and Wertheimer (1938) found
a very substantial reduction in the glycogen reserves in starving
rats. After 3 days fasting, however, the glycogen values rise again.
Similar results were published by Barbour, Chaikoff, ]\'L4.cleod
and Orr (1927), who found in rats higher liver glycogen values
after 48 hours fasting than after 24 hours. These findings are in
complete agreement with the often astonishingly high liver glycogen
values reported by earlier authors, after lengthy fasting tests.
Unlike these authors Lajos (1936) found a continuous fall of
the liver glycogen in rats during 4 days fasting.

Heyman and Modic (1939), in studies on rats of different
ages, found that the liver glycogen content fell with equal rapidity
during starvation, irrespective of the animals’ age. The lov'est
value was reached after 6 — 10 hours.

B. Skeletal muscles. As regards the reaction of the muscle
glycogen to starving, Luchsinger (1875) considered that it
disappeared more rapidly than the liver glycogen. In rabbits
according to this author, it has completely disappeared after
2 days fasting. Aldehoff (1889), on the other hand, found con-
siderable amounts of muscle glycogen in rabbits and cats after
a fast of 6 and 12 — 14 days respectively. IsHiMORl (1913) writes:
»Der Glykogengehalt der Muskulatur zeigt in Uebereinstimmung
mit den bisherigen Erfahrungen keine auffallige Abhangigkeit von
der Nahrungszufuhrs. Junkersdorf (1921) states: »dass der
Muskel im Hungem sein Glykogen, wenn auch in geringeren

Mengen, viel zaher zuriickbehalt als die Leber». Karczag,
Macleod and Orr (1925) found in normal rats a reduction of
the liver glycogen to O .27 % after 24- hours fasting, as compared
with 0.40 % in the controls without fasting. Markowitz (1926)
found that, though starvation tends to lower the muscle and liver
glycogen content, it can never make an animal entirely free from
glycogen. A combination of starvation, cooling and strychnine
spasms can make the liver, heart and skeletal muscles glycogen-free
for several hours. Handowsky (1928) states that the muscle con-
tains an undiminished amount of glycogen after 24 hours fasting.
Lawrence and McCane (1931) found in rats a decrease in the
muscle glycogen from 0.54 % in normal animals to 0.35 % after
24 hours fasting. It then keeps unchanged for the next 24 hours.
Wesselkina (1932) in experiments on cats, found that the muscle
glycogen markedly diminished after fasting. Long and Evans
(1932) found no definite effect on the muscle glycogen of shorter
fasting periods than 48 hours. Dittmar (1933) found in rats, after
a fast of up to 7 days, a slow and approximately uniform diminu-
tion of the muscle glycogen. Agren, Wilander and Jorpes
(1931) found in mice a considerable decrease of the body glycogen
after 10 hours fasting. Heyman and Modic (1939) found the
lowest liver glycogen values after 6 to 10 hours fasting.

Liebig (1940) found in rabbits a decrease of the muscle glycogen
to 0.15 % after 6 — 1 days fasting, as compared with 0.34 % in the
normal animals. Nutter (1941) found a marked decrease in rats
after 24 hours fasting.

Own investigations. For the present investigation the effect of
fasting is of the greatest importance, especially for the studies on
mice. I have therefore made a tabular summary of my analyses
on normal mice with varying fasting periods, all of them decapitat-
ed at 8 a. m. The results are tabulated in Fig. 34 and are graphic-
ally shown in the diagram Fig. 35.

In complete agreement with the results published by Neufeld
and CoLLip (1941), the liver glycogen content in the animals in
my experimental series seems to keep fairly constant during the
first five to sk hours. It then falls rather rapidly and seems to
reach a minimum after about 26 hours fasting, after which it

mice. Females. Decapitated at 8 a.

•to

fej

*•«.»

fe;

C!^ .

.a o c
-j .9 <-
c 't ^'3 ^
c> o ® c>

«^E + +!

§.fe:s1l^

8,.e» §

i? g
O §>

« § CO

-:+l

§,|X

>j §

ns
o
CQ

o SS I 9 >i

“ g clT

— “ _Ll

eo ^ +)

»-< |i J

B «

a ns

o p

H «

l+l

^ ^ ii>

S «

< c

S ,
|>-H
]y.

4 )^

"i+i

lt<!

>»

'tJ

O «
6 .§
C3

O tfi

c
« U3

•« ^

OOt^OO-VO€ 0 <M*f#<eOW« 30 ^

c 4 rHTH 05 T-<c 4 T-Ir-nOrHOOO

+1 +1 +i +1 +! +1 +1 +I +1 +1 +1 +1 +1

l>»^C^COkOCO\Ot»C^C 9 COOtO

CO^>-IcoC<lcio^^t>tDiC 5 t-’c 6

iO CO

O O

kO ••^ ^ CO •-<

00000

CO eo CO

0000000000000

+1 +1 +1 +1 +1 -H +1 +1 +1 +1 -H +1 +1

O I- O <N 04

^ CO CO -j< iCi

000000

0000000000000

I'- »0 — CO 04 W

COOOOCOI'-W

000000

^COCOCOC 4 CQ*^COOOh-r «0
‘04-^C4C0-h— tOOOOO

oooooooooodoc )

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 4-1 4-1 4 l

»0 CO ^ Cfi. C4 00 t- 'TH c^l -?•

cococd 09 Cd(M*- 4 *- 400 «itoi«

cic 5 o 4 c 4 c 3 c 4 c< 5 T-Hddddd

o CO r>-

O rH UO

4 ^ 4-1 4-1 4-1 I 4-1

oj »ft 00

rH CO CD

4 | 4 ^ 4-1 4"1

C»

O'^-**C 0 Ot^-^C 0 C 0 m

'^cococM'^ir^'rtior-to

CO t-
CO 05

4-| 4-1 4-| 4-1 M 4-1 4-) 4-1 4l 4-| 4-1 4l 4-1

963

943

922

T-t

i'*

T-^

730

867

715

056

753

l'-

-H +1 +1 +1 +1 +1 +1

+1 +1 +1 +1 +1

•o

«o

r*4

7—1

05*

00*

l^*

T“4

05 l«- Ol

^ CO O OJ iO 05 05
05 T-i 05 T-i

«5

.9 »»

a ti 4 ^ A M

S ^ pCS

iC 5 rH 01 O CO

M' »0 o CO • 05

>-< 05 05 05 CM CO

Fin. 35. Diagrams shmoing the diminished storage of glycogen in mice
after different periods of fasting. The diajp-ams are based
on the figures tabulated in Fig. 3i.

Per cent

Percent

again rises somewhat. The curves for the body glycogen content
and the total amount of glycogen per Aveight unit show similar
variations. For investigations such as the present one, possi e
changes which set in very rapidly are of the greatest importance,
especially as mice have a very rapid metabolism. I have there
fore examined the magnitude of the glycogen depots in mice 25
minutes after all food had been removed from the cage. As the

variations must be very minute, I concurrently studied a series of
animals without fasting. The results, which are tabulated in Fig.
36, show very good correspondence for the two series.

Summary:

As will be seen from the above account, the glycogen reserves
and especially the liver glycogen, are very labile, and are affected
by several different dietary factors. The experiments which have
been made on this subject are often contradictory and the experi-
mental conditions obscure, whence it is not possible to throw
clear light on these questions. In experiments where the glycogen
reserves are studied it is necessary to work with uniform dietarj
conditions.

It is a generally recognized and accepted fact that starvation
tends to lower the liver glycogen content, and that the reduction
after 24 — ^28 hours fasting is very considerable in all animals
examined. According to the data in the literature, it seems
probable that the liver glycogen content, after reaching a mini-
mum, again rises somewhat. Reports as to how the liver glycogen
behaves after a short fast are sparse. Also as regards the niuscle
glycogen a reduction after fasting is reported, though the figures
are more divergent.

According to my studies on mice, there is no change in the
glycogen reserves during the first 5 — 6 hours of the fast. A steadj
decrease till about 26 hours is then found, and afterwards again
a rise. The glycogen content in the rest of the body behaves in
the same way.

CHAPTER 9.

Cyclic changes in glycogen content anO
yariations due to temperature

A. Diurnal variations. The question regarding variations in
the liver glycogen content at different times of the day has been
the subject of very animated discussion. It has been concerned
not so much with the now accepted view that the liver at different
times of the day contains varying amounts of glycogen as with
the nature and causes of this variation.

The first observations on the subject were published by Fors-
GREN in 1927. Two years later he gave a curve for the diurnal
changes in the glycogen content of rabbit liver. His investigations
were made merely on a small number of rabbits, but they led to a
number of deeper studies with different animals.

The first thorough study on the diurnal variations in the gly-
cogen content was published by Agren, Wilander and JORPES
(1931). Summing up their observations on a large material, those
authors state: i> There are cyclic changes in the glycogen content
of the liver of rats, mice and rabbits, which are to a large extent
independent of the intake of food. Glycogen accumulates in the
liver during the night and disappears again to some extent during
the next morning.*

»Similar periodical changes occur, though to a minor extent, in
the muscles also.*

These results are confirmed, amongst others, by VAN Weel
(1941).

Forsgren (1929) published a cyclic curve with two liver gly-
cogen maxima about 4 p. m. and 2 a. m. and a minimum about
10 a. m. This observation has been confirmed by v. Euler and

Holmquist (1934) and afterwards, on a large material, by Sjo-
gren, Nordenskjold, Holmgren and Mollerstrom (1938).

Holt^iquist (1931) stated that the curve for the liver glycogen
content in the rat ran parallel with that given by Forsgren
(1929) for the rabbit. This statement, however, afterwards proved
to be erroneous. According to Holmgren’s extensive investigat-
ion (1936) on white rats, the liver glycogen maximum for that
animal is at about 8 a. m. and the minimum at 4 — 8 p. m. According
to Deuel, Butts, Hallman, Murray and Blunden (1938), the
liver glycogen maximum for the rat is at about 4 8 a. m. and the

minimum about twelve hours later.

As regards the guineapig, the liver glycogen content, according
to Petren (1939) shows a maximum about 11 a. m. 3 p. m. and
a minimum about 5 — 9 a. m.

According to Seckel and ICato (1938), the diurnal variations
in the liver glycogen content of the rat do not fully manifest them-
selves before the age of 3 weeks. Holmgren (1941) states.
»...dass der Leberrhythmus beim Meerschweinchen schon bald
nach der Geburt auftritt oder eventuell bei derselben bereits vor-
lianden ist».

In regard to the diurnal variations in the glycogen content of
the muscles, the reports are very scanty. Such variations, however,
have been observed by Agren, Wilander and Jorpes.

B. Seasonal variations. The question whether the carbo-
hydrate metabolism is subject to seasonal variations has been much

discussed, whithout, however, being definitely decided. As for
possible seasonal variations in the liver glycogen content, this
question has been studied by many investigators, especially in re-
gard to cold-blooded and hibernating animals. The cold-blooded
animals are of minor interest in this connection, and most of the
studies on hibernating animals are unsatisfactory, being based on
insufficient material.

Gurber (1895), in regard to rabbits, stated that he had found
a liver glycogen content averaging 4.25 % in the summer, as comp
ared with 11.75 % in the winter. He tells us nothing, however,
about the diet, time of day, etc. Kissel (1896), in studies on

rabbits, likewise found a lower glycogen content in the summer
than in the winter. On very loose grounds, he rejects the suppo-
sition that the variation may be due to a difference in food, con-
sidering it to be endogenously involved in the nature of the rab-
bit organism. FuJil (1924), after two years’ study on rabbits at
different seasons, states: »The glycogen content of the liver in 151
rabbits was on an average 3.3 %. It underwent a seasonal varia-
tion: i.e. was definitely smaller in June and July in both the
yearsf (these being the warmest months). CORI and CoRl (1928)
state: »Experiences of the past two years have shown that there
is a great constancy of the preformed glycogen from year to year
as well as at different seasons of the year*. It should be noted,
however, that this statement refers to rats after a fast of 24 — ^28
hours. Burn and Ling (1929) state that the liver glycogen con-
tent in rats on a fat diet keeps rather low in the summer, but is
higher in the autumn, spring and winter. Hirsch and van Pelt
(1937), consider themselves to have shown certain seasonal changes
in the maximum and minimum glycogen content in the liver.
Petrkst (1939) arrives at similar results. Chrometzka and
Beutman (1940) state that the first-mentioned author in 1939
had ascertained, as regards hibernating animals, a higher glycogen
content in the winter than in the summer. As regards the muscle
glycogen, however, he had found the reverse. Goldblatt (1929)
states: »It has been found in these young animals (rats) that the
glycogen content of the liver after 24 hours’ starvation is higher
in' the winter than in the summer*.

As regards the content of muscle glycogen, the reports are scan-
tier. Fujii (1924) found no seasonal variations. Handowsky and
Westphal (1928), in a study on the carbohydrate content in the
skeletal muscles of normal rabbits, state: *Es liessen sich Jahrzeit-
liche Beeinflussungen nachweisen*.

C. The effect of the temperature of the environment. All
those who engage in the standardization of insulin are well aware
that the temperature greatly affects the sensitivity of animals to
insulin, and consider it essential that the temperature should be
kept constant within rather narrow limits. Similar observations
have been made in regard to the glycogen content.

As far back as 1881, Kulz observed that the liver glycogen con-
tent vas reduced by cold. He adopted the rather rough method
of immersing rabbits in icy-cold water, Lanczos (1933 and 1935)
made experiments on similar lines, placing mice for 4 — 5 hours in
a tin placed on ice. Though this did not affect the liver glycogen
content, it rose when the mice recovered in room temperature.
Markowitz (1926) stated that the body glycogen stores were
reduced by cold, Silvette and Britton (1932) that it tended to
lower the content of liver glycogen. Rafferty and MacLachlan
(1941) compared the liver glycogen content in rats which had
been kept at an outdoor temperature of 35.5° — 36.6° C and at
20°— 21° C, respectively. They found a higher liver glycogen
content in the animals which had been kept at a lower tem-
perature, and the difference was statistically significant. BoMSKOW

and V. Kaulla (194'2) state that a low outdoor temperature
entails a 'marked reduction in the liver glycogen values of guinea-
pigs. At an outdoor temperature of 15° C they found merely 38 %
of the liver glycogen content that had been observed at 23° C.
The muscle glycogen was unchanged.

As regards from the above, the statements on this question
show great variation and systematic studies on sufficiently large
material are missing. From the existing scanty reports is seems
probable that the liver glycogen content is affected by the out-
door temperature. Rawson and Guest (1939) consider that a
constant temperature is essential for comparative glycogen in-
vestigations.

Summary ;

Diurnal variations in the liver glycogen content have been
shown in most of our usual laboratory animals. This is of the
greatest importance for all investigations in which different
values for the glycogen content are discussed, and we cannot
avoid reckoning with this factor. An investigation which does not
take into account such variations is valueless. In order to be in a
position to draw any conclusions from differences in glycogen con-
tent between experimental animals and controls, it is essential

that they should have been examined at the same time of the
day. In the present work the experimental series is always comp-
ared with the concurrently examined control series.

The question whether the body glycogen stores are subject to
seasonal variations has not been settled. If seasonal variations
occur, they may also be due to differences in the outdoor tempe-
rature, the composition of the diet, estrus etc. (Handowsky and
Westphal). An actual seasonal rhythm has not been shown.

This question is closely connected with that of the effect of
temperature. To judge by the scanty data in the literature, it
seems probable that variations in the temperature affect the liver
glycogen content.

For the present investigation such seasonal changes and tem-
perature variations are of minor importance, seeing that, as al-
ready pointed out, the experimental animals and controls are
always examined concurrently, and that the temperature in stab-
les and laboratory has been kept between 17° and 22° C.

CHAPTER 10.

Further factors which affect the glycogen
content in normal animals

As appears from the above, there are a number of different fac-
tors that affect the glycogen content in normal animals, and
which can be controlled or avoided in the c.vperimcnts. There arc,
however, other factors which entirely elude control and which one
must try to neutralize in serial experiments.

The reaction of different experimental animals to insulin under
different conditions has been studied very assiduously, and it has
been found that all the factors which, according to the above
account, affect the glycogen stores also affect the sensitivity of
the animals to insulin. A review of the discussion on these quest-
ions is given by v. Ledebur (1936). According to that author,
however, we must also reckon with e. g. climatic influences etc.,
which elude control. Laqueur and de Jongh (1925) speak of
»uberindividuelle Faktoren*. »Damit soil gesagt sein, es kommen
bestimte Tage, auch Perioden mehrere Tage bis zu einer Woche
vor, wo plotzlich sehr starke Reaktionen auftreten konnen, viel-
Icicht auch mehr Krampfe als sonst sich bemerkbar machen, und
andererseits kommen auch Perioden vor, wo nur besonders schwa-
che Reaktionen beobachtet werden. Man muss dann naturlich an
klimatische Faktoren wie Temperatur, Luftdruck, Feuchtigkeit
Rsw. denken.» This fact is well known among all those who work
with standardization of insulin preparations. I have myself ob-
served the same phenomenon.

Similarly, the glycogen depots and especially the liver glycogen
are subject to uncontrollable variations. Even if every precauti-
on is adopted, one may suddenly find a change in the level of the
glycogen content. Such a change is illustrated by the experiments

is a

a <i

•I fe

•W O

^ »

^ 2

r a

^ *v^

?5 S
, «
Q 'S
Oo ^

'1^
8 Oo

Cl §

. <25

•>s

'ta

•T 3

fc.

fe.

?11

•>«i

« S

2 S

ft ^

J? 's

•2

fti

-^|g

g'S'-S «

1 g S s ^ cir

8 ^ .S-a ^

^ C5^‘S c “H
e c rt .

^ »- -w *x

8 P'.'=» s

2 ? g § §

O -g IX

O g CO

•a;:-H

^§.i«

o' ti IX
„ 2 CO

^^+\

S g.l«

& cu
5 w '

C ns
^ c

e §

,i>^
e w

l-H

^\i<

SP IX

I S '«

t- g

l-H

l«

6.1

CO T-»

+1 +1

to ^

CO 05

O O

+1 +1

o o

+! +1

»-4 CD

+1 "H

-s<

O CO
OJ CO

to CO

CO 05

-H -H

^ CO
CO nfi

o o

-fl -H

iTi

■r^

CO*

<£>

tabulated in Fig. 37. Both groups of animals were treated in exact-
ly the same Avay. The experiments were made Avith an interval of
one Aveek. In each group half of the animals received O .004 and
half O .008 TJ of insulin subcutaneously Avithout preceding fast.
They Avere; decapitated at 8 a. m., one hour after the injection. The
difference beUvcen the tAVo groups is greater than the difference
within the groups, that is, betAveen the animals Avhich had received
different doses. As Avill be seen from the table there is a statistic-
ally significant difference betAveen the two groups in regard to
the glycogen content. As regards the body glycogen there is
a good correspondence. This is thus apparently the result
of the so called »uberindividuelle Variations’. If the intestinal
AA'eights in the tAA'o groups are compared, it Avill be found that the
alimentary canal of the animals richer in glycogen is heavier than
that of those poorer in glycogen. The difference, Avhich is 640 ± 129
mg is statistically significant. It is thus conceivable that the differ-
ence in the glycogen content may be due to the possibility that
the former animals during the night had eaten more or later than
the latter and consequently had a higher liver glycogen content.
The reason for this possible change in the intake of food, hoAvever,
eludes control. These sudden variations, Avhich occur also in regard
to the glycogen content in normal animals, entail certain conseq-
uences in judging the results. Obviously, Ave cannot immediately
compare tAVo othenvise exactly equivalent groups of animals
examined at different times. In order that the values may be com-
parable, the studies must have extended over a considerable length
of time, so that the oscillations dn different directions can equalize
one another.

By comparing an experimental series with a concurrent control
series, this error, of course, can be most surely avoided. The pos-
sible effects of this kind can be obviated by extending such com-
parisons over several days. This has been done throughout in the
present investigation.

PART III

THE EFFECT OF EXOGENOUS
INSULIN ON THE GLYCOGEN STORAGE
OF NORMAL ANIMALS

CHAPTER 11.

Survey of tlie reports in the literature on the
effect of insulin on the glycogen storage
of normal animals

A. Different Avaya of studying the effect of insulin on
the glycogen storage.

Attempts have been made in different ways to obtain an idea
of the effect of insulin on the glycogen depots in normal and
diabetic animals. Here only the studies on normal animals will be
dealt with.

1. In vitro exiperim-ents with tissues and tissue slices. The effect
of insulin on the tissues has been studied in vitro. See, for example,
Ahlgren (1925), V. Euler (1930) and Blaschko (1933).
Attempts have also been made to study the effect of insulin on
the tissue glycogen content in vitro. The results have been rather
meagre and contradictory.

Thus, for example, Collazo, Handel and Rubino (1924') state
»...dass der Glykogenabbau in vitro im Leberbrei nicht in nen-
nenswerter Weisse beeinflusst wird dutch Insulinzusatz.» Seckel
(1938) states: »With a method described in an earlier publication
rat liver glycogenolysis as it normally occurs in surviving tissue
slices suspended in a buffered salt solution, Avas shown to be
inhibited to a considerable extent by large doses of insulin added
in vitro->.

^Because of this finding and the results reported in the experi-
mental and clinical literature, the essential action of insulin on
the liver is believed to be an inhibition of the glycogenolytic
process particularly when the latter is proceeding at a high rate.*

Stadie, Lxjkens and Zapp jr (1939) state: jNo effect upon
carbohydrate synthesis by liver slices was observed. The carbo-
hydrate synthesis by diabetic cat liver slices was not found
increased; here, too, insulin was without effect .» The same authors
emphasize this point of view in a publication of 19410: »The new
formation of carbohydrate by liver slices of normal or diabetic
animals was found to be uninfluenced by the addition of insulin
to the equilibrating medium*.

In experiments with rat diaphragms in a Warburg apparatus with
a method indicated by Gemill 1940, Gemill, and Hamman
(1941) state that they had observed a storage of glycogen in the
muscles when Insulin was added to the glucose containing suspen-
sion medium. The increase in glycogen was small and the reported
experiments were few in number, but the results are regular and
uniform.

Hechter, Levine and Soskin (1941) work with the same,
methods. They state: *Glycogen deposition in rat diaphragm in
vitro varies directly with the concentration of glucose in the
medium. Insulin catalyzes this process greatly at low sugar con-
centrations (100 mg %), but very little at high concentrations
(400 mg %).»

Summing up, it may be stated that attempts to study the effect
of insulin on the glycogen storage in vitro have not led to any
results as regards the liver, but seem to indicate that insulin under
certain conditions may stimulate the storage of muscle glycogen.
The absence of effect in experiments in vitro is, of course, no
proof of a similar condition in experiments on animals, as the
metabolism of the tissue is radically changed in experiments
in vitro.

2. Perfusion experiments on surviving organs, for the study of
the liver metabolism, were proposed as far back as 1853 by
Claude Bernard. The technique in regard to tortoises was
indicated by Grube (1907). This animal has the great advantage
that both liver lobes can be perfused separately, in which case
one of them can serve as a control. In a number of experiments
it was shown that, in perfusion tests with the liver of cold-blooded

animals, the building-up of glycogen could be observed. See
review by e. g. Kapfhammer (1925) and Neubauer (1925).

It is unnecessary to review here all the different perfusion
experiments that have been made on cold-blooded animals. It need
only be mentioned that Kepinov (1938) had found that unpurified
insulin preparations increase the glycogenolysis in perfused frog
liver, which was not the case with crystalline insulin. This observa-
tion is of the greatest importance, especially as most of the perfu-
sion experiments were made at the beginning of the insulin era.
At that time investigators worked with more or less impure pre-
parations, which greatly reduces the value of their results. It has
been found very difficult to develop perfusion technique so that
it could be applied also to the liver of hot-blooded animals, which
is very sensitive. The perfusion was at first made with Ringer
solution, and the glycogenolysis could not be mastered. The earlier
literature on the subject has been reviewed by de Meyer (1909)
and Barrenscheen (1913). The latter, however, states that sbei
geeigneten Versuchbedingungen gelingt es, auch an den iiberleben-
den isolierten Warmbliiterlebern mit regelmassigheit Olykogen-
ansatz zu erzielen*. Laufberger (1924) states: »Bei Durch-
spiilung der iiberlebenden Hundeleber trat trots Insulin eine Ver-
mehrung des Zuckers in der Durchblutungsfliissigkeit ein».

Bernard (1925) perfuses rat liver with Ringer solution with
the addition of glucose in different concentrations. The liver
retains sugar proportionally to the glucose concentration in the
perfusion liquid. This retention is doubled on the addition of
insulin. The retained sugar, however, is not stored in the form
of glycogen. jEine Glykogensynthese \vurde mit unserer Methodik
nicht erzielt. Das Insulin war ohne Einfluss auf die Geschwindig-
keit der Glykogenhydrolyse.»

The results of perfusion experiments on organs from hot-
blooded animals are not very encouraging, and CoRi (1931) sum-
marizes his experiences as follows: »The results obtained on the
isolated mammalian liver are not conclusive . . . But the chief
difficulty in experiments of this kind seems to be that it is practic-
ally impossible, even with the best technique available, to keep
the mammalian liver outside of the body for any length of time
Without some serious loss of normal function. For this reason an

effect of insulin which occurs in the intact animal, may fail of
demonstration in the isolated organ.*

Fiessinger et alios (1937, 1939 and 194D) state, however,
that, with a perfected technique, they had been able to perfuse
monkey liver without obtaining any increase in the glucose con-
tent of the perfusion liquid, that is to say, without finding any
increase in the glycogenolysis in experiments which may have
continued for a couple of hours. When they supplied insulin to the
perfusion liquid, the liver gave off considerable amounts of glucose
to the blood for a time, but soon recovered equilibrium despite
the fact that it still contained considerable amounts of glycogen
and, on a new supply of insulin could again give off large amounts
of glucose. Similar results were reported by Bodo and Marks
(1928): * Glycogen storage in the absence of added insulin was
observ'cd . . . Added insulin either stopped further storage or
caused a breakdown of glycogen.* The differences reported by
those authors were, however, very small. Corey and Britton
(1941) found, on perfusion of cat liver, that insulin tended to
reduce the glycogen content. On the supply of suprarenal extract,
they observed an increase.

As regards muscles, the perfusion method has not been of so
much interest. Here, instead, investigators could use animals from
which the liver had been extirpated for similar studies. (Mann
and co-workers. Best, Hoet and Marks, 1926, etc.) In a sum-
mary, Mann (1927) states that the supply of insulin increases
the consumption of glucose in the muscles. He considers that the
liver has no bearing on the development of insulin hypoglycemia,
but only on the restoration of the normal blood sugar level. Best
et alios likewise consider that insulin acts only via the muscles.
When glucose in the blood is in- excess, glycogen accumulates in
the muscles. When there is a shortage of glucose, the already
existing glycogen is retained.

Summing up, it can be said that perfusion experiments with a
view to studying the effect of insulin on the glycogen metabolism
have been concentrated on experiments with liver. The results
have not given any support to the view that insulin stimulates
a storage of glycogen in the liver. On the contrary, Fiessinger
and Bodo and Marks indicate that insulin causes an outflow

of glycogen from the liver. Other experiments pointing in the same
direction are devoid of conclusiveness in view of evident defects
in the technique.

3. Attempts to draw conclusions regarding the efect on the
liver glycogen jrom the variation of the blood sugar after insulin
injection. Mann and IVIagatii (1924) have given a summary of
their excellent studies on the part played by the liver in the
carbohydrate metabolism, from which it is quite evident that the
liver is the only source of the blood sugar. It should be noted,
however, that the blood sugar level is dependent on two different
faetors: the outflow of glucose from the liver and the absorption
of glucose in the tissues. For this reason it is impossible to draw
any conclusions from the rise or fall of the blood sugar level with
regard to an increase or decrease of the glycogenolysis or possible
glycogenosis in the liver. We may quite as well be concerned with
a decrease or an increase in the peripheral consumption, while the
conditions in the liver remain constant. It is still more impossible
to draw from the variations in the blood sugar any conclusions
regarding changes in the glycogen stores of the liver. The liver
glycogen, as we know, is a very labile factor. In regard to its
amount, it is at any given moment dependent on the relation
between glycogenosis and glycogenolysis, a matter Avhich is often
overlooked. Thus, under certain conditions we may find an increase
of glycogen concurrently with hyperglycemia (adrenaline effect at
certain stages, according to CORI and others).

A rather naive view — which has been put forward quite seri-
ously — , is that the blood sugar, when it disappears under the
effect of insulin, is deposited as liver glycogen and causes a rise
in the liver glycogen content. On the contrary, the liver during
the whole time when the insulin is acting must, of course, give
off glucose to the blood even if the blood sugar level is low, and
perhaps particularly just then. Under the action of insulin, the
combustion of glucose is generally considered to be stimulated
and the amount of glucose in the blood will, of course, suffice
merely to maintain the combustion for a few minutes. Fresh
glucose must be continuously supplied and it can only come from
the liver.

From the above it should be evident that all discussions about
the behaviour of the glycogen depots under the action of insulin
on the basis solely of blood-sugar studies are meaningless, and I
consider it therefore unnecessary to cite any literature on the
subject, even- if it is very abundant.

4. Studies on the effect of insulin on the glycogen stores in
intact animals. In studying the effect of insulin on the glycogen
stores by direct analyses of organs, investigators have proceeded
in two quite different ways, a) Attempts have been made to
determine the glycogen content of the organs before and after
the supply of insulin and b) the glycogen content in animals after
the action of insulin has been determined and compared with the
corresponding value for untreated animals.

a) Brentaito (1939), takes as a, basis the following statement
by Geelmuyden 3>es ist unmoglich bei Kaninchenversuchen
selbst nach der peinlich uniformen Vorbehandlung Kontrolltiere
mit einigermassen gleichem Gehalt an Leberglykogen herzustellen.
Wenn man . . . entscheiden will, ob Insulin die Glykogenspeiche-
rung . . , fordert oder hemmt, so sollte diese Entscheidung auf
Grund eines Vergleichs zwischen dem Glykogengehalt des Tieres
vor und nach der Insulingabe getroffen werden. Diese Ideal-
forderung is aber in der Praxis kaum zu erfullen.> Brentano
continues; »Diese Forderung Geelmxjydens haben wir — wenn
ich nicht irre, als erste — in dera vorliegenden Arbeit erfiillt*.
Brentano works with normal rabbits. Under light ether
anethesia he extracts a muscle from one of the hind legs and a
piece of liver for glycogen analysis. After a couple of days the
same procedure is repeated after treatment and supply of insulin.
Here, however, there are various possibilities of error, including
the so-called »uberindividuelle» variation (LaqueuR and DE
Jongh).

Another jmssible proceduce is, in the course of the experiment,
to take samples from the liver and muscles and in this way con-
tinuously to follow the glycogen variation. Most investigators who
have tried this procedure have had their laboratory animals under
narcosis. In this way, however, they exposed themselves to the
complicating action on the glycogen which may be attributed

directly or indirectly to the narcosis. Coni (19.31) states: »\Ylien
this is done under anesthesia, one must be aware of the fact that
all anesthetics so far available, including aniyLal, have a depressive
influence on glycogen formation in the liver. Furthermore, anes-
thesia and laparotomy, even if the greatest possible care is taken,
arc liable to be followed by an increased release of epinephrine
from the adrenals.* This method lias been adopted especiallj'^ by
French investigators and in work with larger laboratory anim.als,
where for reasons of cost it is not possible to arrange extensive
series.

Attempts to obtain continuous samples without placing the
animals under narcosis have also been made. Mouxoii and Poi/-
liACK (1930) laid a lobe of the liver of rabbits and dogs outside
the abdominal wall in accordance with a method indicated by
Kijnz and Molitoii (1929), after which they could take their
samples. The method entails many sources of error and has not
been used by others. A more promising experiment was made by
Com and Com (192.3), who constructed an abdominal window
with a detachable lid, to be sewed into the abdominal Avail of
rabbits and dogs. A few days after this operation the experiment
IS made. The detachable part of the windoAV is screwed off and
samples can be repeatedly taken from the liver. The greatest
druAvback of this method is the tendency to hemorrhage.
Moreover, the manipulations Avith the liver seem to lead to a
continuous decrease in liver glycogen, also in control animals.
The method Avas adopted and tested by several inA'estigators,
but Avas subjected to rather a scA'crc criticism by scA'eral authors,
including Greatsnstuic and Laqueur (1925 and 1926), Enms-
mann (1927) and ITandowsky (1928). Com himself soon
abandoned his method, and in 1931 he Avrote: »A number of
authors haA'e used this method AA'ith moderate success. The
difficulty AA'as that stopping the hemorrhage by cautery, Avhilc
very effective, proved to be an unsatisfactory procedure because
the animal is severely damaged by absorption of toxic products.
Recently it has been found that a high frequency current applied
to the cut surface of the liver Avhilc it is being compressed stops
the hemorrhage perfectly Avithout causing more than a superficial

Fig. 38. Survey of the most important and most often cited papers

Author

Animal

species

Time of
fasting

Quantity of
insulin

Time between
injection and
estimation

Dudley and Marrian (1923)

Mice

—

1 hr

Con C, F. (1925)

15-60'

V. Meyenburg (1924')

—

Low and Krcma (1929)

0 — 16 hrs

0.6 U

l-l-J- hr

CorHl (1930)

18-24 hrs

—

Tonutti and Wallraff (193S)

2 hrs

Barbour, Chaikoff, Madcod

Rats

24 hrs

lU/kg

1 hr

and Orr (1927)

24 hrs

2U/kg

1 hr

24 hrs

3U/kg

1 hr

Goldblatt (1929)

24 hrs

0.2-0.6 U

—

Low and Krcma (1929)

0 — 16 hrs

6 U

1 — 1-j hrs

Lamcnce and McKanc (1931)

24 hrs

0.1 U/100 gm

2 hrs

Daoud and Gohar (1933)

—

Russel (1938)

24 hrs

1— lOU
per kg

44 hrs

Spitzbarth (1940)

24 hrs

5 U

—

Goldblatt (1929)-

—

Goldblatt (1933)

Guinea

pis

24 hrs

1— 2U

Collazo, Handel and Rubino

■»

18 hrs

7 U

4 hrs

(1924.)

Spitzbarth (1940)

24 hrs

4 U

—

BabJdn (1923)

Rabbits

—

Dudley and Marrian (1923)

—

6 hrs

100

on the effect of inxiiUn on the glijcogen storage of the intact animal.

Effect on liver
Slycogcn

Effect on muscle
glycogen

Num1>cr
of control
animals

Numl>cr
of insulin
animals

Decrease

—

Dccapit. at first
convulsion

Decrease

—

If)

No change

Histological c.xaminnlion.
Practically no glycogen
in the livers of control
animals ciUicr

Decrease

IJecrcn.se or no

—

IJccapitatcd at Ijcginning

change

of convulsions

No change

Decrease

—

No glycogen in
the liver

—

Histological method

Decrease

Large decrease

—

SevemI animals.

» »

Different treatment

No change or

> >

—

slightly decreased
Decrease

Decrense or no
change

Slight decrease

•Almost complete

•Considerable

• Convulsive doses*

depletion*

diminution*

Decrease

Increase

The animals are given
glucose simultaneously
with insulin injection

l-Mge decrease

Increase

—

Food ad lib. until decap

No change

—

Certain animals

Increase

—

Glucose simultaneously
with insulin injection

Decrease

•Marhcdly

lowered*

•Markedly

lowered*

Decrease

Large decrease

Decapitation at first
convulsion

101

Author

Animal

species

Time of
fasting

Quantity of
insulin

Time between
injection and
estimation

Brugsch, Benatt, Horsters aud

Rabbits

24 hrs

Kata (1924)

Hcymans and Heymans (1925)

—

Bonn (1925)

24 hrs

50 U

2 — 3 hrs

Frank, Nothmann and Hart-

3-6 d.

0.1 TJ/kg

3 — 4 hrs

mann (1925 and 1927)

V. Meyenhurg (1924)

24 hrs

—

4- -5 hrs

Grevcnstuk and Laqucr (1925)

4 d.

O.l U/kg

2 hrs

Fisco (1926)

5 d.

—

Markowitz (1926)

5-10 d.

—

Ehrismann (1927)

1—4 d.

—

HandowsJcy (192S)

0—24 hrs

Varying

—

Rossi (1928)

5—10 d.

5 U.

—

Sakyun and Luck (1929)

24: hrs

25 U/kg

1 hr

Goldblatt (1929)

24 hrs

0.6 U

1—3 hrs

24 hrs

0.2 U

1—3 hrs

24 hrs

0.5 U

1 — 3 hrs

Goldblatt (1930)

48 hrs

0.2—1 U

2—5 hrs

CorMl (1930)

24 hrs

0.6 U

—

Loch, Nichols and Paige (1931)

0-2 d.

7—75 U/

kg i. V.

—

CorkUl, Marks and White

24 hrs

—

(1933)

102

Ettecl on liver
glycogen

Effect on muscle
glycogen

Number
of control
animals

Number
of insulin
animals

Large decrease

in convulsive
stale

Decrease

—

Convulsions thought to

No cliange

No change

3 + 3

give decrease in liver
glycogen

•Convulsive doses*

Considerable

No change

Blood sigar lonered. No

increase

Considerable

oonviilsions

Histological examination

increase

Sometimes

increase,

sometimes

decrease

—

Abdominal svindow

Large increase

Several small insulin

> »

injections during fasting

Several animals.

No increase

Different treatment

Abdominal window

—

Decrease

d = 2f

Oisappeared

Disappeared

• ..gclbtct sobald die

No chang;c

No change

Insulin-wirkung ilue
Ilblicpunkt crrciclit
liatte*

Considerable

No certain

10 veeks old

increase

Considerable

changes

No certain

c » *

increase

Considerable

changes

Decrease

C » » . Severe

increase

Increase

No change

convulsions

Considerable

Increase

Decapitated as soon as

increase

Decrease

flaccid and coordination
lost

Increase

103

Author

Animal

species

Time of
fasting

Quantity of
insulin

Time between
injection and
e.slimntion

Goldblatt (1933)

Rabbits

24 hrs

1 U/700 gm

Sunaba (1936)

24 hrs

0.5-2 U/kg

—

Bridge (1938)

12 hrs

1-4 D/kg

6 hrs

Brentano (1939)

•

24 hrs

5 D/kg

5 hrs

Goldblatt (1930)

Cat

2-5 d.

0.2—1 D

2 — 5 hrs

Reid (1936)

48 hrs

—

2 — 3 hrs

Fisher and Laclcey (1925)

Dog

Bhri^ann (1927)

24 hrs

—

2 hrs

Biirger and Kohl (1935)

0.6 U/kg

—12 hrs

Vendeg (1935)

0.06—0.1 U

—

0.1- 0.6 D

0.6-1 D/kg

Hebb (1937)

12-48 hrs

—

Corkill (1930)

Ferrets

24 hrs

12 D

3 hrs

Rathery, Giberl and Laurent
(1930)

Monkey

12 hrs

repeated
small doses

17 — 25 hrs

Rathery, Gibert and 1/iurent
(1931)

—

0.76 D

—

Christol, Jiedon, Loubaticres and
Monnier (1938)

repealed
small doses

—

3 D/kg

104

Effect on liver
glycogen

’

Effect on muscle
glycogen

Increase

Slight decrease

Decrease

Increase

No change

No certain change

—

Decrease

No increase

. -

Number
of insulin
animals

Decrease

Slight increase
Decrease

Considerable

decrease

Decrease

No change
Decrease

Increase

Decrease

No change

Slight in

Sampling in narcosis

Intravcnuous infusion of
insulin and glucose

Glucose simultaneously

Infusion of insulin
intravcnuously,

0.07 — 0.28 U/kg and hr

Abdominal window

Sampling in pemoclon
anaesthesia

Sampling under pemocton
anaesthesia. Insulin and
glucose injected intraven

Given insulin and
dextrose. Sampling under
anaesthesia

Sampling under
anaesthesia

Glucose simultaneously

Sampling under
anaesthesia

Sampling under
anae^hesia

105

necrosis. Even with this improvement serial determinations on
rats seem preferable to taking samples of liver from one and the
same animal.®

B. Experiments with series of insulin-treated animals
and control animals

have proved to be the working method which has yielded
the best results in studying the effect of insulin on the gly-
cogen depots. Also here, however, there are many lurking
sources of error (see the immediately preceding section of this
work). In particular, investigators do not seem to have reali."ed
the magnitude of the biological variation and have accordingly
worked with too small series. The biological variation, howe'fer,
is of fundamental importance. LAQtiEtrR and de Jongh (1925)
clearly explain the immense importance of this factor. They
state: ». ..ergibt sich aus dem obigen Material die Sinnlosigkeit,
aus ein oder drei Versuchen Schlusse zu ziehen. . . . Denn, wenn
sich herausstellt, dass die ’Streuungsbreite’ eine sehr bedeutende
ist, dann diirfen einerseits XJntersuchungen an einem oder drei
Tieren nich gemacht werden, andererseits, wenn dies noch ge-
schieht, brauchen sie. jedenfalls nicht mehr gelesen werden.® These
points of view, as may be gathered from the following summary,
have unfortunately been entirely disregarded by a large number
of investigators.

In works on this subject statistical treatment of the material
is often non-existing, so that the authors have no clear idea as
to what conclusions their experiments entitle them to draw. In
some extremely rare cases the conclusions drawn are not suf-
ficiently far-reaching, but in most publications the reverse is
the fact.

In Fig. 38 the principal and most frequently cited of the papers
on the effect of insulin on the glycogen stoves, made by direct
analysis of organs in experiments on intact normal animals, are
tabulated.

Very thorough investigations into the effect of insulin on the
body glycogen have been made by CORI and CORI and their
co-workers. With the aid of the previously mentioned abdominal

106

window, CoRi, CoRi, and Pucker (1923) studied the effeets o
insulin on the liver glycogen of a rabbit and arrived at the result
that »Iletin causes glycogen synthesis during ingestion of glucose
even though the blood sugar and free liver sugar is below that
of a normal starved animaU. These experiments, however, were
few in number and not quite convincing. C. F. CoRi (1925) states,
as a result of studies on rabbits with the abdominal window
method: »Our data on the influence of insulin on the liver glycogen
of starving rabbits show that there is no appreciable change
the glycogen or total carbohydrate content within the irst ou
of insulin action, whether the initial glycogen content is
low, or whether the fall in blood sugar is slight or strong, n
second to sixth hour of insulin action the glycogen content o t le
liver may remain constant or may decrease.* He also examine
mice which had received O.oi — 0.04 U. of insulin. It was oun
that they reacted much quicker, for which reason all the analyses
were made 15 — 60 minutes after the insulin injection. »The aver
glycogen content of the liver of the injected mice was 39 pe
cent lower than the average of the corresponding controls, n on y
one half of these experiments did a decrease in the glycogen
content of the liver of the injected mice occur, while the other
half showed no change in the liver glycogen after the injection
of insulin.* In regard to the animals which had concurrent y re
ceived glucose, he writes; >Iii summarizing this section t e
important fact seems to be that the insulin-injected anima
only synthesizes glycogen in the liver, when it has an excess o
sugar available, but that the rate of glycogen deposition is actua y
increased*. Normal fasting animals who receive insu in, on e
other hand, show no increase in the liver glycogen content, w
is interpreted to be due to the fact that they have no
able for storage in the form of glycogen. *It is conclu e
insulin produces glycogen synthesis whenever there is a c

excess of sugar available.* ,

The abdominal window method was later aban one

severe criticism by several authors. . ,

CoRi (1926) and CoRi and CoBi (1926 and 1928) worked wit
rats which had fasted for 24— 4« hours. They received g ucose
through a tube in an amount which guaranted resorption in a eas

107

four hours. These investigators computed the combustion, the
amount of glucose resorbed and the amount of glycogen deposited.
»The average amount of sugar absorbed was nearly the same in
both series.* The animals which had received insulin deposit less
glycogen in the liver and more in the rest of the body than the
controls. CORI and CORI (1928) supplement their above reported
experiments on the effect of insulin in the »absorbtive state*
with investigations in the >postabsorbtive state*. After 24 hours
fasting the rats receive through a tube, per 100 g. of weight,
2.5 cc of a glucose solution containing l.OGS mg glucose. This is the
amount of glucose which on an average is resorbed in four hours
under the said conditions. Four hours after feeding, the insulin
injection experiments begin. The amount of glycogen which the
rats at this time should show is estimated in accordance with
previous experiments. The changes in the course of three hours
are determined. *The insulinized rats utilized three times more
liver glycogen but only slightly more body glycogen than the
control rats . . . According to this analysis, insulin is a hormone
that leads to a preferential utilization of blood sugar and indirectly
of liver glycogen, the latter being the only important source of
blood sugar in the body.* CoRl and CoRi (1929) determined the
effect of insulin on rats concurrently with intravenous infusion
of glucose in large amounts. *For an equal quantity of glucose
injected, the insulinized animals showed a blood sugar level and
a glycogen content of the liver approximately one half of that
of the control animals. When a similar blood sugar level was
maintained in control and insulinized animals by injecting less
glucose into the former than into the latter there was no differ-
ence in the amount of glycogen deposited in the two groups of
animals.* CORI and CORI (1929), after experiments on rats which
had received insulin after fasting, conclude: *Die Insulinwirkung
ist am Hungertier nicht sehr Eindrucksvoll, was bei dem geringen
Glykogenbestand der Leber nicht zu verwundern ist. Denn Insulin
erhoht die Kohlehydratverbrennung vor allem auf Kosten des
Leberglykogens, und zwar in der Weise, dass infolge der gesteiger-
ten Aufnahme von Blutzucker in den pcripheren Geweben mehr
Zucker von der Leber nachgeliefert wird. Muskelglykogen wird
erst dann in verstarktem Masse angegriffen, wenn der Glykogen-

108

bestand der Leber niedrig ist.» CoRi, CoRl and Buchwald
(1930), in regard to non-fasting rats which had received 0.24 units
per 100 g, state: »15 minutes after the insulin injection liver
glycogen and blood sugar were unchanged, while after one hour
both had diminished!'.

Summary.

As appears from the above account, there is a considerable
divergence of opinion regarding the effect of insulin on the
glycogen storage. Most of the investigations were made on very
few animals and are therefore, in view of the marked individual
variations, not very convincing, even if they sometimes perhaps
contain correct observations. But also the investigations which
have been made on a larger number of animals have given con-
tradictory results. Thus, for example, it is stated in regard to
the liver that insulin entails an increase in the glycogen storage
(Frank, Nothmann and Hartmann; Goldblatt; Corkilb;
Vendeg;) whereas a decrease is reported by CORI and CoRi,
Lawrence and McCane, Russel, Bridge, Burger and Kohl,
Vend^:g. As regards the muscles an increase was noted by
Russel, Corkill, Bridge and Cori and Cori, no change by
Frank, Nothmann and Hartmann, Goldblatt, whilst e.g.
Handowsky as well as Lawrence and McCane found that
insulin tended to lower the muscle glycogen content.

It is difficult to say to what these contradictory results may
have been due, but they may partly be explained by the fact
that the experimental conditions were not comparable. As a rule
merely the effect of a single dosage was examined and the result
of it was analyzed some time after the injection. Merely sporadic
attempts to study different dosages under the same conditions
were made by Vendeg, Goldblatt, Cori and Cori. These
experiments, however, never comprised sufficiently varying dosages
and they were never extended to analyses at different times after
the injection, except in experiments on narcotized animals where
samples of the liver were repeatedly taken during the test.

It is impossible by studies of the literature to arrive at a
reliable view regarding the effect of insulin in excess on the
glycogen storage of normal animals.

109

CHAPTER 12.

Effect of insulin on the glycogen
stores in mice

Most of the experiments were made on mice, as these animals
had shown the greatest uniformity in their reaction. This had
also the great advantage that the whole animal can be used for
analysis.

The experiments were divided into four groups; 1. The mice
received the insulin injection without preceding fast. After the
injection they were starved until decapitation. 2. The mice receiv-
ed the insulin injection after a fast of 24 hours and were after-
wards starved until decapitation. 3. The mice received the injec-
tion after 24 hours fasting and then had free access to food.
4. In some of the mice such small doses of insulin were injected
that the blood sugar was not affected. The planning of the experi-
ments is indicated by the table in Fig. 39.

The experiments in group 1. were made in order to study the
effect of insulin on as >normal» an organism as possible; those
in group 2. in order to study the effect on an organism with
reduced glycogen content and without a supply of carbohydrates
in the intestinal canal. In the literature it has in fact been con-
tended that the effect of insulin on the glycogen depots is to a
certain extent dependent on the- supply of carbohydrates. The
experiments in group 4 were made in order to control the state-
ments of Frank, Nothmann and ILvRTHtANN (1925 and 1927)
and others that insulin in very small doses entails a storage of
glycogen in the liver.

The condition of the mice at the time of decapitation can be
roughly described as follows; 1. Unaffected. 2. slightly

110

Fig. 39. Table showing the 'planning of the experiments cm normal mice. The table gives the treatment, the insulin
doses and, the time of insulin injection. All animals decapitated at 8 a. m.

affected: apathetic, sit still and cowering in a corner, but
not paralyzed. 3. Greatly affected; more or less paralyz-
ed, unconscious, considerable fall of temperature, spasms.

Group 1. Those doses were chosen because a dosage of O .004
U. merely caused a slight reduction of the blood sugar without
affecting the general condition, whereas the dosage 0.2 U. affect-
ed it in a very marked degree. This latter dosage, however, was
not so large that it led to death, at any rate not within the time
covered by the experiments.

None of the mice died from these insulin doses. With a dosage
of O.004 XJ. insulin subcutaneously all the animals remain through-
out completely unaffected. With a dosage of O.oos U. some of the
mice were slightly affected after two hours. After the lapse of
four hours or longer all of them seemed to be unaffected. The
dosage O .02 U. slightly affects some of the mice after one hour,
markedly affects a few of them after two hours and slightly
affects the majority of them. A few of the mice still remain
unaffected. After four hours all the mice appear to be lively and
unaffected. The dosage O .2 U. after one hour already affects all
the mice, most of them slightly, a few markedly. After two and
four hours they are more strongly affected. Eight hours after the
injection some of the mice are still slightly affected, whilst the
others are unaffected. Such a short time as 25 minutes after the
injection all the mice are unaffected by all the dosages.

The general condition of these animals is thus most affected
2 — 4 hours after the insulin injection. The effect sets in earlier and
continues longer, the larger the dose is.

The variations in the liver glycogen are shown in the tables
Figs. 40 — 49 and in the diagram Fig. 50. The diagram shows clear-
ly that all the insulin dosages employed, give reduction of
the liver glycogen content. 25 minutes after the injection
the tendency to reduction is noticeable as regards all the do-
sages and after 1 — ^2 hours it reaches a maximum. Afterwards
the liver glycogen content again begins to rise. The tables show
that the reduction of the liver glycogen content 1 hour after the
injection is so marked that the difference between the controls
and the insulinized animals is statistically significant for all the

112

114

iir»

Fig. JfG. Table. Normal viice. Females. Food ad lib. No food after insulin injection at a. ni. Decapitation at S a. m.

Glycogen content in
mgm per 10 grn of body
weight minus alim.
canal and spleen

X ± £-

-it \0 CO 00 Cl

03 th T-t d d

+i +1 +1 -H -H

Cl vO oo o

CC5 d d CO d

CM i-f »H tH

Body glycogen
per cent
X±E-

00 <C OO 1'*. 00

o o o o c>

o o o o o

d d d d d

+1 +1 +1 -H +1

CO CS CO Cl 00

CO CO t'- O CO

o o o o o

d d d d d

t£ .

S g cl,'^

CO «a « o CO

-?• Cl Cl w o

d d d d d

-H +1 -H +1 +1

Via o CO •-<

Cl «o i-w o CO

03* CM -H d d

s ^ ivi

a Ji ^ i
. I'-H
^ix

< e

o CO i^a

tH ITS C3 O

+1 -H -H -H +1

O Ci O

CM C3 ifl C3

03 CO C<l

<03 C<l 03 CM 03

Liver weight
mgm

x±.e-

954 ± 28

993 ± 38

958 + 43

879 ± 28

917 ±28

Body weight
gm

X±E-

20.9 ± 6.5
22.2 ±0.8
21.3 ±0.5
21.2 ±0.5
21.7 ±0.6

No. of
animals

03 03 03 03 03
v-^ rK 1-i 1-1

Dose

Controls

O.OOl

0.008

O.02

0.2

•S

•w

t5>

si;

S 2 c

s e-S

.£:s “

c w 5
<y 2 ®

o ^ S
5

cj^

.'T3

-H

co^

■H

a o
O O

o o

V V

W C3

03 CM

»o O

•-« <-1 o «-•
« 00
03 id lo

00 Cl VO «o
03 03* 03 03

-H +1 -H -H

-?< 00 »o «
CO cd ^ cd

*-< o o o
O O O O

l*- w l"*

o o d d

03 03 CM 03
01 CM CM CM

o O O iO

o VO O vO
vO OS "ij

t-1 d r-» d

O O O O

o o d d
+1 +1 +1 +1

^ o c« o
o o o o

d d d d

^ o

I I ®

VO VO ^ ^

d s V V

03 03 03 03
03 03 03 03

O 00 O O
O 00 O l-
1- « Cl O

d 03 d d

O Cl vO CO
vO

d d d d

+1 +1 +1 +1

vfi t-*. -j*

CO »-• €0 «C
d t-J 03* 03

o O O CJ

d d d d

116

Fig. 4S. Table. Normal mice. Females. Food ad lib. No food after huvilin injection at 12 p. rn. Decapitation at S a. m.

117

Fig. 50. Diagrams showing the effect of different insulin doses on blood
sugar and storage of glycogen at different lengths of time after injection.
Food ad lib. After insulin injection no food. All animals decapitated
at 8 a. TO. All values given in per cent of the
corresponding values of the controls.

Percent

118

dosages; likewise 2 hours after the injection. 4 hours after the
injection only the two largest dosages show a statistically sig-
nificant difference as compared with the controls, whilst the differ-
ence for the dosage O.oos U. must be designated as probable
(P = 0.05 — O.02). In the animals which had received the smallest
dosage, O.004 U., the liver glycogen content by that time had risen
so much that it showed a good correspondence with that of the
controls. 8 hours after the injection all the animals except those
which had received the largest dosage showed good correspondence
with the controls. As regards the animals which had received a
dosage of O.2 U., a reduction is still statistically probable (P =
= 0.02 — 0 . 01 ).

Variations in the body glycogen content: The diagram shows a
slight reduction of the body glycogen content for all dosages 1
hour after the injection. Aftenvards we note as regards the dosage
0.2 U. a marked rise, so that these animals 2 hours after the injec-
tion show a body glycogen content exceeding that of the controls.
The variations after the other dosages are less pronounced. It is
evident from the tables that the insulinized animals 25 minutes
after the injection showed a good correspondence with the controls.

1 hour after the injection there is a difference bordering on pro-
bability (P = 0.05 — 0.01) between the controls and the animals
which had received O.oos U. 2 hours after the injection there is no
convincing difference between any of the insulin series and the
controls. If the body glycogen content in the animals which had
received O.2U. and in those which had received O.02 U. is compar-
ed, we find a difference of O.040 % ± 0,oi4 (t = 2 . 857 ; df = 31; P =
= 0.01 — O.001), which must be regarded as statistically very prob-
able, if not certain. 4 hours after the injection there is no statis-
tical difference for any of the dosages as compared with the con-
trol animals. 8 hours after the injection the animals which had
received the largest dosage showed a statistically probable in-
crease as compared with the controls (P = 0.05 — O.02). The ani-
mals which had received the other insulin dosages show good cor-
respondence with the controls.

Variations in the total axmount oj glycogen: It will be seen from
the diagram that all the dosages employed result in a reduction
of the body glycogen stores. The diminution culminates after

119

1 — 4 hours and then slows down, so that all the animals 8 hours
after the injection show a good correspondence with the controls.
One hour after the injection the diminution is statistically sig-
nificant as regards all the dosages; likewise after 2 hours.

As regards the bhod sugar (cf the diagram) one can observe
as regards all the dosages a falling tendency 25 minutes after the
injection. After the lapse of one hour the blood sugar level has
again begun to rise after the two smallest dosages, whereas it still
falls after the two larger ones. Four hours after the injection the
blood sugar level in the insulinized animals is again about normal
except in those which had received the largest dosage, 0.2 U.,
in which it is still lowered. Eight hours after the injection even
these animals begin to show a normal blood sugar content.

Recapitulation: The effect of insulin on the general condition
of the animals and on the blood sugar level is proportional to the
dosage. It sets in more rapidly and continues longer, the larger
the dosage.

All the insulin dosages tend to lower the liver glycogen content.
The largest doses give an increase of the body glycogen content.
The total amount of glycogen per wieght unit of the animal
diminishes under the effect of insulin.

Group 2. The results of these investigations have been tabulated
in Figs. 51 — 58 and are graphically shown in the diagram Fig. 59.

Effect on the general condition: One hour after the injection
all the mice seemed to be still unaffected. Two hours after the
injection most of those which had received the dosage O.02 U.
are slightly affected, whilst those which had received O.2 U. are
greatly affected. After 4 and 8 hours respectively a few of the
mice are slightly affected by these larger dosages. After the
dosages O.004 and O.oos U. they appear to be unaffected.

The variations in the liver glycogen, as sho^vn by the diagram,
arc essentially similar to that of the animal in the preceding
experimental group. It will be seen from the tables that the reduc-
tion one hour after the injection must be regarded as statistically
probable for the dosage O.004 U. (P = O.os — O.02) and as statistic-
ally significant for the other dosages. The situation is the same

120

Fig. 51. Table. Normal mice. Females. 24 hrs. fasting. No food after insulin injection at 7 a. m. Decapitation at 8 a.

Glycogen content in
mgm per 10 gm of body
weight minus alim.
canal and spleen

r4 O O O O

-H +1 +1 +1 +I

C5 *'t

la o lO o

Body glycogen
per cent

x±e-

i5 O ^

O O O O O

o o o o o

d o d d

+! 11 +1 +1 +1 -H

!>• — eo « c» w

c5 o rt CO o •?*

o o o o o

O 0 0*0 O*

Liver glycogen
per cent

X ± £-

0.67 ± O.OG

0.40 ±0.05

0.86 ± 0.03

0.30 ± 0.03

0.26 ± 0.02

i J

E T5 ®

WS p r’

< «

0 o oo o

CM Ci O 00

rH tH

+! +1 +1 +1 +1

cj to en CO 05

r- C5 00 o —

01 i-i eo W

C'J W « CJ

Liver weight
mgm

x±e-

lO 00 CM

o o CO

+1 +1 -H -H +1

r* CD ir5 — * CO

CD CD iTi O O

OO 00 CO

Body weight
gm

X ± £-

rM O 00 kO 1'*

yA O O O

+1 +1 +1 +1 +1

S-* I'- 00

00 00 c**’ r-*

rH rH tH

No. of
animals

CM O CM CM CM

T -1 tH rH rH T-f

Dose

Controb

0.004

0.008

O.02

0.2

121

Fig. 53, Table. Normal mice. Females. 24 krs. jasling. No food after instdin injection at 0 a. m. Decapitation at S a.

Glycogen content in
mgm per 10 gjii of body
weight minus niini.
canal and spleen

t£> -4 ^ CO Ci

o ^ o ® ^ ^

-H II -hI +1 +1 -H

00 ^ ^ Cl o o

Body glycogen
per cent

X + e-

eo ei »-• CO «

o o o o o

O o O O O

-H +1 +1 -H -H

CO CO CO -rt »o

o o o o o

O o O O o'

Liver glycogen
per cent

X ± £-

<X3 iO -4 O CO

o y— s, o ✓-V o o o

o^o^o o o'

-H II +i II -H +1 -H

— O Cl I'- »o

o* o' O* O O

"5 c
eg

< «

lf5 O O O CJ

O'! CO to CQ

M +1 +] +1 -H

ift o O O? CO

O *-< o

t- CO oi cn CO

tH 1-4 03 1-4 tS

Liver weight
mgm

x±e~

to 00 to t- 1 -

W T-< T-< T-4

M +1 -H +1 +1

O o lC5 O 4

■— < 1—4 lO lO t-

t— 00 t>

•SP \>i

O Cu

i 1*^

18.3 ± 0.2

18.7 ± 0.2

20.4 ± 0.5
19.0 ± 0.3
18.3 ± 0.2

No. ot
animals

o r- to

1-4 ^ tM ^ T-4

Dose

Controls

0.004

0.008

0.02

0.2

122

Fig. 55. Table. Normal mice. Females. Si hrs. fasting. No food after insidin injection at a. m. Decapitation at 8 a.

123

124

two hours after the injection, when the reduction is maximal as
regards all the dosages. Four hours after the injection we find a
statistically significant reduction only as regards the largest
dosage, whilst the mice which had received the other dosages now
show a good agreement with the controls. Eight hours after the
injection only the animals which liad received the largest dosage
show a significant reduction of the liver glycogen content.

Variations in the body glycogen content: As shown by the
tables, we find in the mice which had received 0,2 U, an increase
in the body glj'cogcn content which is statistically very probable
or certain two hours after the injection, and must be regarded as
probable four hours after the injection (P = 0.05 — O. 03 ). As appears
from the diagram, the tendency is the same one hour after the
injection, and still remains so after eight hours. As regards these
two latter points, however, the increase is not statistically certain
in the present material. The reaction to the smaller dosages is
less uniform and the values for the body glycogen content in the
insulinized animals show throughout good correspondence with
those of the controls.

Variations in the total amount of glycogen: As regards the gly-
cogen content per weight unit, we always find a good correspon-
dence between the insulinized animals and the controls, e.xccpt
in one respect: the mice which had received O.oos U. show 2 hours
after the injection a reduction as compared with the controls
(P = 0 . 0 .^. — 0 . 02 ).

VariatioTis in the blood sugar: As regards all the dosages we find
a diminution of the blood sugar, most marked 1 — 2 hours after the
injection. The blood sugar then again rises, gradually approaching
the normal value. The diminution is greater and continues longer,
the larger the dosage (cf. the diagram Fig. 59).

Recapitulation: In this experimental group the effect of insulin
on the general condition of the animals and on the blood sugar
level is proportional to the dosage. It sets in more rapidly and con-
tinues longer, the larger the dosage.

All the insulin dosages tend to lower the liver glycogen content.
As regards the variation in the body glycogen content, we find an
essential difference in effect between the largest dosage and the

125

Ftg. 59. Diagrams showing the effect of increasing insulin doses on blood-
sugar and storage of glycogen in mice at different times after injection.
After 24 hrs. fastiny insulin injection, then food. Decapitated at 8 a. vu
All values given in per cent of the corresponding
values of the controls.

Peretni

126

smaller ones, in that the latter do not result in any certain change
in the body glycogen content, whereas the dosage 0^ U. does.

Group 3. The results are tabuLaled in Figs. GO — G5 and are illu-
strated in the diagram Figs. G6 and 67.

All the animals appeared to be unaffected.

Variations in the liver glycogen: The control animals, which
after a 24 hours’ fast, were given free access to food, showed
during the first 8 hours’ a very substantial continuous in-
crease of the liver glycogen content. If they are compared mth
those which w’cre decapitated at 8 a. m. after a fast of 24 hours,
it •will be found that the liver glycogen content has increased
2.5 times in one hour and 7 times in four liours. If the controls are
compared witli the animals which had received the different insu-
lin dosages, we nowhere find any statistically significant difference.
As regards the animals which had received the dosages 1 and 20
U., one hour after the injection there is possibly a slight decrease
as compared with the controls (P = 0.05 — O .02 and O .05 respectively).
. If we study tlie diagrams in Figs. GG and G7, it will be found
that the animals which had received the larger insulin dosages
0.2 — 20 U., one liour after the injection showed a tendency to
diminution of the liver glycogen content which, in magnitude and
duration, was proportional to the dose.

Variations in the body glycogen content: In the controls, the
body glycogen content shows a marked continuous increase when
the animals after 24 hours fasting arc given free access to food.
Those which had received the smaller insulin dosages show
throughout good correspondence with the controls. Those which
had received 0.2 — 20 U. show 8 hours after the injection an in-
crease in the body glycogen content as compared witli the normal
animals. In those which had received 0.2 U. insulin, the increase
is so marked that it is statistically probable (P = 0.05 — O. 02 ). In
those which had received 1 U. it is very probable or certain, and
in those which had received 20 U. it is statistically significant. At
other times after the injection the insulinized animals show,
statistically, a good correspondence with the controls. From
the diagram, however, we already see a tendency to an increase
in the body glycogen content 4 hours after the injection of the
two largest doses.

127

Fig. GO. Tabic. Normal mice. Females. 2-1 hrs. fasting. Food ad lib. after instdin infection at 7 a. in.

Decapitated at S a. m.

e’l* .

^ w Ci *

H te 3 w Co

8 '='.£'3 ,|

gi{::S'3

g R..531 g

±>g i g
o <>> '-
s

s §

•s^+i

o ti ix

« § CO

U ^ \iA

a a

Sis
I'S i«

-< ej

a g CO

2 '"iX

.£P IX
V Co
A

O o

O O

+i +{

t-4 C)

c-l

op O

M II -H

-H -H

kO ^ ct

r-? C4 03 c4

+) +1 -H +1 +!

^ lO O 00

O '! T -< 1 — <

o c o o o

0*0 0 0 0

-H -H -H -H +i

« th 00 CO ta

eo o

0 0 0 0 * 0 *

o o o o o

-H -H -H -H -H

t'' C» CO W w

05 03

O W O
krd 00 o

+1 +i +1 -H "H

Cfi

G<i

0^3

-H

■H

-H

-?<

€0

»o

o’

-H

■H

mfl

«?

T-I

T-<

T*«

Yl ©1 d

128

(

GSl

0.2

0.02

0.004

H-

CiT’

i<9

»—

>-*

<r>

•‘i

•‘I

•^1

♦— *

o»

H-

»—

►«•

>-*

•^1

-‘j

•»!

)-k

•^1

H-

-“1

►-*

J—

»S^

c»

va O

H ^

^ S ft
B ^ S
§ 'St g

£3 tt ft

5 B

0*3

5 *2

B B

Fig. 61 . Table showing the difference between controls and insuUnized animals tabidatcd in Fig. 60 .

Fig. 62. Table. Normal mice. Females. 34 lirs. fasting. Food ad lib. after insulin injection at 4

Decapitated at S a. m.

Glycogen content in
mgm per 10 gm of body
weight minus tilim.
canal and spleen

— 1 «£> M

C<i OiJ

■H -H +I +1

e-j e?5 w

c6 CO -ji ^

lA iO If) MO

CM ^ I'* eo

do CO CO CO*

+17+1+1+1

«~l O kO

r-J vS CO d

CO CO CO lA

® \X

^ ,1
“k +i

O O xo

eo v-«

o o o o

O O O O

+1 -H -H +1

O O Cd o

00 03 t>> CO

C4 C« d

d o o o

•V «>4 O

04 CQ ^

o o o o

d odd

-H +1 -H +1

%a C5 • xo

Ci *-♦ kft

C4 eo CO e3

d d d d

Liver glycogen
per cent
x±.e-

1-- tv

CO n

iS cS cS cS

-H +1 4^ +1

vH CO xO

xo eo ct CO

d xA d \o

e> ^ xf3 -n

x£3 «}<

d 2 CO

+1 j| +1 +1 M

eo rt c'S •-< Ci

tx> >0 O 04 C5

d CO

'3 a

n « M

a ^ Co

^ &f+l
I"?

lA I-* CO CO

rH »-( T-i tH

*f| -fl -f) -l-j

<fi CO to

^ CD tv

O fv tv O

CO Ol

v}< Csj

O t- CO CO

ri CM rn t-4

-f| +1 "H +1

O tA CO i-<

O O CO A?

04 04 CM O

CO CO CO 07

.'5' ty,

w ^ Cg
i ^ 1

^ s ,

fN CO CO

+! +1 “H "H

r- CO o CO

lA CO CO CO

O QO CO CO

CO 04 C4 lA

rjl ^ T-1 CO

+1 +1+1+1
o CO lA CO

CO OJ TV CO

O O O C3

r-< tH TV

^3 jx

o (g

•n |X
o '

d d d d

+1 +i +1 +1

CJ C5 <0 CO

d d d d

Ol r-* T-f T-!

o xo -*< TJ*

t- 3 odd

+1 +1+1+1

cs o o

CO d cd cd

rH 1-t TV

No. of
animals

CO CO CO CO

TM r-* T-< T-l

tV rV tV t-v

rv rv T-< T-r

Dose

.2 f «

g o o

i: o o o

g o d o

«n

1 ei o

= O rH O

130

Fig. 63. Table showing the difference between controls and insulinized animals tabulated in Fig. 62.

131

Fi/j. r>.'f. Table. Normal mice. Females. Itrs. fastinff. Food j,d lib. after insulin injection at 12 p. m.

Decapitated at S a. m.

c’S .

H c

C5 4!

c S 3 « clT'

8 '='.3'^ Ij

^ o s c -H

d o

■^'7

-H

l>-

-H

g ^ “ I J

C“»

-*<

c6«S

t'-

8 R-.? 5

8 s

O ~

»A

t£

o ti IX

se s ««

— S jj
to +1

-H

O '

I-

c»

C-J

l’«.

«pH

•!<

o ^ X
t> R cii

^ _i )

6£ “H

-Hii

•H

u ^ |>>

-c

*^*

4>H

*« c

c f; iH

T-»

o ^ « «u

O £» fc .

. ^ ^ +

-H

-M

I'-

< <3

«i2

*«4

.Ef ,!«

U .4 Co

u ^+f

■H

£!

iCS

»Q

-tH

»o

■I'

.Ef IK

u C4l

^ l+l

-H

■H

c»

t-.

■g

CO*

1-^

o C5

• H

>-<

T-^

T-l

^ c
<4 cs

132

Fig. 65. Table shoioing the difference between controls and instdinized animals tabidated in Fig. 6h.

133

Fig. CG. Diagrams showing the effect of insulin on blood-sugar and storage
of glycogen in mice. Insidin infection after ZJf hrs. fasting; then food a d
lib. All values given in per cent of the corresponding values of the con-
trols, which after 24 hrs. fasting were given their food simultaneously
with the visidin animals. Decapitated at S a. m.

Percent

SHl

2 ♦ < t irs

Blood sugar.

Liver glycogen content.

Body glycogen content.

Fig. 67. Diagrams showing the effect of time on the effect of increasing
doses of insulin on blood sugar and glycogen storage in mice.

The same experiments as given in Fig. 66.

Percent

Blood sugar.

Liver glycogen content.
Body glycogen content.

135

Variations in the total amount oj glycogen: The amount of gly-
cogen per weight unit in the insulinized animals shows throughout
good correspondence with the controls.

Variations in the blood sugar: As regards the dosages 0.2 — 20
U., we find a decrease of the blood sugar 1 hour after the injec-
tion. As for the two smaller dosages, the blood sugar then rises to
the normal value, whilst as regards the largest dosage it remains
low during the whole period of observation. It is remarkable that
the blood sugar curve and the liver glycogen curve always run
parallel when there is a marked change in the liver glycogen con-
tent (cf. Fig. 66 and 67).

Recapitulation. Mice which after 24 hours fasting are given free
access to food show during the next 8 hours a substantial conti-
nuous increase m the glycogen stores. The supply of small
amounts of insulin has no effect on this process. The supply of
larger amounts of insulin promotes the storage of glycogen else-
where than in the liver, and the largest insulin dosage tends to
reduce the storage of liver glycogen. It is noteworthy that the
diminution in the storage of liver glycogen runs parallel with a
sinking of the blood sugar level.

Group 4. It has been stated in several quarters that the quali-
tative effect of insulin on the glycogen stores is due to the
amount of the dosage, and that by varying the latter it would be
possible to obtain an increase or decrease in the glycogen con-
tent of the liver. Investigators have used the terms »physiolo-
gical» and »non-physiologicaU dosages. By ^physiological dosage*
they apparently mean a dosage which does not lower the blood
sugar level (Frank, Nothmann and Hartmann, 1925 and 1927,
Staub, 1930, and others).

In the experiments in group 3, no change in the blood sugar
level is found after the smaller insulin doses (cf. the diagram in
Fig. 66). Thus, though an effective insulin dosage had been given
to these animals while the blood sugar level was high, no storage
of glycogen beyond that of the controls was obtained.

As these tests, however, were not entirely comparable with
those of the above-mentioned authors, further investigations were
instituted. They were made on mice which had received injections

130

137

Fig. 70. Tabic. Normal mice. Females. hrs. fasting. Subcutaneous saline
injection at 7 a. m. No food. Decapitation at 8 a. m.

Dose

No. of animals

Liver
glycogen
per cent

Body
glycogen
per cent

Glycogen con-
tent in mgm
per 10 gm body
•weight minus
alim. canal and
spleen

Blood sugar
per cent

Controk

0.32 i 0.04

0.044 ±0.003

5.6 ± 0.4

0.088± 0.002
(n-8)

NnCl

0.37 +.0.08

0.040 ±0.002

5.6 ±0.3 .

0.08C ±0.002
(n = 10)

O.05 ±0.06

0.004 ±0.004

O.l ±0.6

of O.ooi and O .002 U. without previous fasting. After the injection
they were left without food until the decapitation at 8 a. m., one
hour after the injection. The results are tabulated in Fig. G8 and
G9. As regards none of these dosages could any change in the gly-
cogen content be observed and the smallest dosage, as Avas ascer-
tained, did not entail any change in the blood sugar level.

It is, of course, conceiA-able that in the case of these small insu-
lin dosages, the irritation involved in the actual injection might
entail changes in the glycogen stores which might be mistaken for
the effect of the insulin. True that the well-knoAvn »Fesselungs-
hyperglykami* Avas avoided, but it has been shoAvn that emotional
e.vcitation may cause an increase in the blood sugar (for the lite-
rature, see Holmgren and "Wohlfahrt As Mann and

his co-AA’orkers have convincingly shoAA-n that Ave need not reckon
Avith other sources of the blood sugar than the liver glycogen, it
must be concluded that a change in the liver glycogen content had
actually occurred, unless, concurrently Avith the rise in the blood
sugar, the injection had stimulated the glycogenosis of the liver.
SiLVETTE and Britton (1932), in connection Avith a study of the
glycogen storage, state; ^Profound emotional excitation for a
brief period also brings about change in carbohydrate A’alucs es-
sentially similar to those observed after severe exercise.*

138

In order to settle this matter, I arranged serial tests with fasting
mice. I injected a corresponding amount of physiological saline
solution subcutaneously and after one hour analyzed the glycogen
content in the mice. They were then compared with a concurrently
analyzed control series of fasting mice. The results are tabulated
in Fig. 70. It will be seen from the table that the two series show
good correspondence throughout, which indicates that the injec-
tion as such had no effect on the amount of the glycogen stores.

Recapitulation. These experiments show that with very small
insulin doses, if they affect the liver glycogen content at all, always
gives a reduction. There was no indication in support of the view
that insulin in very small doses, so small that they do not affect
the blood sugar level, brings about an increase in the liver glycogen
content.

139

CHAPTER 13.

Experiments on rats

Very thorough studies on the effect of insulin on the glycogen
stores of rats have been made by CORI and CoRi; These studies
were as a rule made on a large material and the results are con-
vincing. On most points thej' correspond with my experiments on
mice. For this reason merely a few experiments were made with
rats, namely where the results diverge and in a few other interest-
ing cases.

Group 1. According to CoRI and CoRl (1929) insulin does not
affect the liver glycogen content in rats which receive the injection
after a fast of 24 — 48 hours. Under these conditions the muscle
glycogen is reduced. As these results differ considerably from my
own observations in the experiments on mice, a control investi-
gation was made. After a fast of 24 hours the rats received in-
sulin subcutaneously, were afterwards fasted for two more hours,
after which they were decapitated and analyzed. The results are ta-
bulated in Fig. 71 and 72. As the results show, CORl and CoRl’s
statements that the liver glycogen under these conditions is not affec-
ted by the supply of insulin were completely confirmed. Also as
regards the muscle glycogen, my results correspond with the figu-
res given by CoRl and CORI, As regards the smaller dosage there
is no change, but for the larger one a statistically probable reduc-
tion (P = 0 . 0 o — 0.02). Wc arc concerned here, however, with very
small figures. Moreover, most of the animals which received the
larger dose of insulin had convulsions, for which reason no im-
portance can be attached to the reduction of the muscle glycogen
content. The glycogen content of the skin in these experiments
shows no marked variations. The insulin dosages were so adjusted
that the smaller ones had left the animals completely unaffected

140

141

Fig. 73. Table. Normal rats. 27/. hrs. fasting. Food ad lib. after insulin injection, at 1 p. m. Decapitation at 5 p. m.

Blood sugar
per cent

C5 1-^ «5 C5
v-i cj

^ o

o o o d

Skin glycogen
per cent

X±£-

r-J

O O O O

d d o d
+1 +1 +! -H

»Q C3

CJ C5 <N

d d d d

Muscle glycogen
per cent

X ± e-

wa <o \o -*i
o o o o
o d d d

-H +1 +1 +1

00 t'- O O
t> CO CO o

d d d t-J

Liver glycogen
per cent

x±e-

3.72 ± O.30
3.39 ± 0.8-1
3.18 ±0.28
2.17 ± 0.38

Liver weight
gm

iO -xf* ^ eo

d d d d
+1 -H +1 -H

00 to t'-

iO d d in

Body weight
gm

x±e~

148 ±13

143 ± 10

13G ± 10

152 ±10

No. of
animals

Cl a Cl Cl

Dose

Controls

0.1

(i.

0.1 — 0.05

0.02 -O.Ol

C<J

O I'*

<N (N

Cl c-

o ©

•0

d d

+1 +1

^ a

. T3

o •-

d d

1 1

w o

d d

PLh

! 1

o =;

<N Oi

O CO

*i>

T-<

d d

w I-

-a

d d

-H

M +1

d Cl

o Cl

d d

<=> d

1 1

^■. s

O d

(N Cl

•V

CN (M

CO o

t***

rA Ci

wmt

d d

-H -H

to Cl

r-< O

142

at the time of decapitation, whilst the larger doses at the time of
decapitation had greatly affected half of the animals and had
slightly affected the other half.

Group 2. In these experiments, the rats after a 24 - hours fast
were given an insulin injection and afterwards free access to food.
In addition to the liver and muscle glycogen, the glycogen con-
tents of the skin was also determined, in order to control statements
made by previous authors such as Scoz (1929) and Richter
(1931), who stated that insulin under these conditions had not
affected the storage of glycogen in the fatty tissue and the skin.
The results are tabulated in Fig. 73 and 74. After the two smallest
dosages O.i and 1 U. per 100 gm of body weight the blood sugar
level remains unchanged. But after the largest dosage 10 U. to 100
gm, the fall in the blood sugar is slight, but noticeable. As regards
the smaller dosages, all the glycogen analyses show good correspon-
dence between the insulinized animals and the controls. As regards
the largest insulin dosage we find a statistically probable reduc-
tion of the liver glycogen content, an almost certain increase of
the muscle glycogen content and a probable increase of the skin
glycogen content (P = O .02 — O.oi). These investigations show tiiat
insulin under the experimental condition in question does not sti-
mulate the storage of glycogen in the liver and that lai^e doses
counteract it. These results correspond to CoRi and CoRi’s in-
vestigations into the effect' of insulin during »the absorptive statei..
As regards the muscle glycogen, the glycogen storage is stimulat-
ed by large doses of insulin.

Summary.

In complete agreement with CORI and CoRi, insulin under the
experimental conditions in question cannot be shown to have any
effect on the liver glycogen content of fasting animals which re-
ceive no food after the injection. The smaller dose does not seem
to entail any change in the muscle glycogen content.

In fasting animals which receive food after the injection the
large insulin dosage entails a reduction of the liver glycogen sto-
rage and an increase of the glycogen storage in muscles and skin.

143

CHAPTER 14.

Exiieriiiieiits on ratobits

The most confusing and contradictory statements regarding the
effects of insulin are found in that part of the literature ^vhich
deals with investigations on rabbit? (cf. the surv'ey of the litera-
ture in Fig. 38). It has therefore seemed to be necessarj' to investi-
gate the effect of insulin also on these animals.

Group 1. The rabbits received an injection of insulin at 3 p. m.
without a preceding fast. After the injection thej' had no access
to food. They were decapitated and analyzed after the lapse of
three hours. The results are tabulated in Fig. 75. According to
these investigations, insulin under these conditions has no effect
on the glycogen stores of the rabbit, at any rate not with the do-
sages employed, 10 U. per kg of body weight subcutaneously, and
at the time when the analyses were made.

It is a generally known and recognized fact that the glycogen
depots in the rabbit are subject to very great variations. This was
the case also in my experimental series. In this way a minor effect
of the insulin might possibly be disguised. Moreover, the alimen-
tary canal of the rabbit contains a large and very variable amount
of carbohydrates, which may affect the results. In view of these

Fig. 75. Table. Normal rabbits. Food ad lib. No food after iiisulin
injection at 3 p. m. Decapitated at G p. m.

Dose

Number

animals

Body weight
gm

Liver weight

<7771

Liver
glycogen
per cent

Muscle
glycogen
per cent

Controls

1938 ± 74

74.8 i 4.1

9.28 ± 0.82

0.88 ± 0.04

10 U kg

1908 ± 106

69.7 i 4.C

8.27 ±0.9G

0.40 ± 0.08

144

Fig. 76. Table. Normal rabbits. 2Jf hrs. fasting. No food after insulin
injection at 3 p. m. Decapitated at 5 p. m.

Dose

Number
of animals

Body weight
gm

Liver

weight

Liver
glycogen
per cent

Muscle
glycogen
per cent

Blood
sugar
per cent

Controls .

2231 ± 77

58.8 ± 7.0

0.2 U/kg

2162 + 61

63.4 ± 3.0

1 U/kg .

2319 ± 118

59.4 ± 3.9

0.23 ±,0.04

marked individual variations in the glycogen stores, the rabbit
must be considered to be an unsuitable experimental object for
studies of this kind.

Group 2. According to several authors, expecially Goldblatt
and CoRKiliL, an increase in the liver glycogen of fasting rabbits
is found on the supply of insulin. This entirely deviates from what
most investigators have shown in regard to other kinds of animals
as well as from my own results in experiments on mice and rats.
A control investigation therefore seemed to be indicated. In my
experiments I used the dosages 0.2 and 1 U. per kg of body weight.
When the animals were decapitated two hours after the injection,
almost all of them seemed to be unaffected. The results of the gly-
cogen analyses are tabulated inFig. 76. As shown by this table, the
correspondence between the insulinized animals and the controls is
throughout good. It is, however, also evident that the individual
variations in the liver glycogen content even in fasting animals is
very marked. With a view to further control of this important
question, serial experiments were arranged as follows. The rabbits
after a 24* hour fast were given 0.5 U. insulin subcutaneously per
kg of body weight, at 1 p. m. As soon as they had become
apathetic, and had difficulty in keeping an upright position, they
were decapitated and analyzed. A control animal was always
examined concurrently with every insulinized animal. The results
are tabulated in Fig. 77. By computing the mean difference and
its standard error in this experimental series, a systematic differ-
ence in the liver glycogen content will be found between insuliniz-
ed animals and controls, the liver glycogen content of the

145

Fig. 77. Tabic shoicing the effect of insulin on glycogen storage in normal
male rabbits, bortrs fa.sting. No food after subcutaneiis injection of O .5 V
of instdin -per Icgm. of body -weight at 1 p. m. hmdinhed animal and cor-
responding control animal decapitated as soon as the insxdinized animal
shoiccd flaccidity and loss of coordination.

Insulin animals

Control animals

•— "a

0 £

w JL

C S2

U •'** t*

'S,

Body weight
gm

Liver weight
gm

Liver glycogen
per cent

Body weight
gm

Liver weight
gm

Liver glycogen
per cent

Liver glycogen. Me
insulin and control
per cent

Liver glycogen. DiR
between control and
lin animal per a

Difference in
per cent of mean

14.25

III

0.25

1650

38.8

0.42

0.84

0.17

14.30

42.1

0.88

ft «

42.5

0.78

0.83

-O.io

-12

14.35

H Si

36.4

0.26

V 0

45,2

0.37

0.32

0.11

14.50

1 '

27.0

0.60

ft W

42.2

2.99

1.76

2.49

140

15.00

ft iS

43.7

1.06

w S

58.2

2.82

1.94

1.76

15.16

|kSi

46.0

0.36

51.6

1.80

0.83

0.94

113

15.50

ftSS

32.9

O.co

ft m

39.0

1.22

0.91

0.62

16.15

1400

40.7

0.66

ft S

54 8

0.88

0.77

0.22

16.20

•nMu

50.4

1.85

tSSiS

70.9

3.71

2.78

1.86

16.25

40.1

2.32

1 750

46.7

0.90

1.61

-1.42

-91

d ± €5 - 49.8 ± 10.1
t = 4.981
df= 9

P = < O.OOl

insulinized animals being lower than that of the controls. Thus
even rabbits react by a reduction of the liver glycogen content
under these experimental conditions, thereby showing no essential
difference in their reaction from other laboratory animals com-
monly employed. It is difficult to find out the reason why authors
such as GoldbIiATT and CORKiLt. have arrived at opposite results.
A conceivable possibility is that they have not taken into con-
sideration the source of error involved in the diurnal variations
in the liver glycogen content. Whether this is actually the fact
cannot, however, be inferred from the scanty data in their
publications.

146

Summary.

In view of the marked individual variations to which the glyc-
ogen storage in rabbits is subject, these animals are unsuitable
for investigations of this nature.

The results of the investigation do not bear out Goldblatt’s
and Corkill’s view that insulin injections into rabbits bring
about a storage of glycogen in the liver, but rather indicate a
reduction, as in the case of other experimental animals.

147

General suryey of results

According to the general view, the liver and the skeletal muscles
are the principal storehouses in the body jor glycogen. Only under
certain special conditions, free access to food after fasting for
some length of time, can the adipose tissue and the skin also store
considerable amounts of glycogen (Chapter 4).

In Chapter 5 the distribution of glycogen within a glycogen
depot is discussed. The statements in the literature vary. The
author’s own investigations do not rule out a difference in glycogen
content in different parts of the liver, but they argue against a
systematic difference of this kind, such as that a certain lobe of
the liver should regularly contain more glycogen than the other
parts of the liver. As regards the muscles, we find marked varia-
tions in glycogen content between different muscles from the same
animal as well as between the same muscles from different animals.
A statistical analysis of the results, however, provides no basis for
any systematic difference. Thus in serial analyses, a result from
one muscle is considered to be representative of the entire
musculature, as also a piece of liver for the whole liver.

The mode of killing and the postmortal glycogenolysis (Chapter
6) has also been considered to play a large part in the amount of
glycogen that could be shown in the depots. It has been stated
by various investigators that the samples should be taken during
a rapidly induced narcosis, but, on comparison, it has been found
that this method has no advantages as compared with decapita-
tion. Nor does freezing in liquid air afford any advantages. The
investigation into postmortal glycogenolysis shows that we need
not reckon with it as a complicating factor if the pieces of tissue
are plunged into hot KOH within two minutes after decapita-
tion.

The age and sex of the animals are considered by a number
of authors to affect the glycogen storages in the body (Chapter 7).
The statements in the literature are based in several cases on

148

large serial experiments. In my own investigations, however, no
sex differences could be observed. On the other hand, it appears
that the age of the animals might have some bearing on the liver
glycogen content. No effect on the muscle glycogen content has
been observed.

The effect of the composition of the diet and the length of the
fasting period is discussed in Chapter 8. From the literature,
various investigations indicating that the composition of the diet
is of great importance for the amount of the glycogen storage
have been cited. The length of the fasting period is also an import-
ant factor. A summary of investigations on animals with different
fasting periods shows that both the liver and the muscle glycogen
content is greatly dependent on the length of the fasting period.
The reduction of the glycogen content in mice culminates after
about 24 hours fasting, and the glycogen content then shows a
tendency to rise again.

Data from the literature regarding variations in the amount
of the glycogen stores according to the time of day, the season
and the outdoor temperature are summarized in Chapter 9.
Diurnal changes must be considered to occur, likewise variations
due to the temperature. On the other hand, no certain seasonal
variations due solely to the season seem to have been ascertained.

In addition to these factors, which can be more or less con-
trolled or eliminated by the experimenter, the glycogen depots
are affected by other factors which are entirely beyond control
(Chapter 10). These factors can be neutralized by concurrently
examining experimental animals and controls, by distributing the
serial experiment over several days and by subjecting the results
to statistical analysis.

There are a large number of reports on different factors which
affect the sensitivity of the individual to insulin. All these factors
have proved to be the same as those which affect the storage of
glycogen in normal animals.

Chapter 11 contains a short review of the different methods of
studying the effect of insulin on the glycogen depots and of the
different results attained by experiments on intact animals. As
indicated by this survey, no agreement has been reached in
regard to the effect of insulin on the glycogen storage. Most of

149

the investigations have been made merely on very few animals,
and, in vicAv of the marked individual variations, are therefore
bj' no means convincing. But even those investigations that have
been made on the basis of a larger material have yielded con-
tradictor}' results. Thus, for example, Frai^k, Nothisxann and
Hartmakxst, Goldblatt and Corkill state that insulin tends
to increase the storage of glycogen in the liver, whilst CORI and
CoRi, Lawrence and McCane, Russel, Bridge, Burger and
Kohl consider that they can show a decrease. Vendeg considers
that the result depends on the amount of the dose. As regards
the muscles, an increase of the glycogen content after the supply
of insulin is noted by Russel, Corkill, Bridge, Cori and CoRi,
a decrease by e. g. Laavrencb and McCane, and by Handowsky,
whilst Frank, Nothmann and Hartmann as well as Gold-
BLATT fail to find any change.

The experiments recorded in the literature were as a rule made
with single dosages and the results were analyzed at a certain
time after the injection. Merely sporadic attempts were made,
for example by Vendeg, Goldblatt, Cori and Cori, to compare
different dosages. These comparative tests were never systematiz-
ed, nor have the conditions at different times alter the injection
been investigated to the desirable extent.

Chapter 12 reports a systematic investigation on mice into the
effect oj different insulin dosages on the glycogen storage at differ-
ent times after the injection and tinder different food conditions.
On each occasion the investigations were made on several mice
and the results were subjected to statistical analysis. It appears
from these studies that the supply of insulin brings about a reduc-
tion of the liver glycogen content. This applies to all dosages that
have any effect at all. The decrease in liver glycogen, in amount
and duration, is proportional to the dose. No increase of the liver
glycogen storage after the injection of insulin has ever been
observed. Furthermore, the interesting fact was noted that the
reduction in the storage of liver glycogen always runs parallel with
a reduction in the blood sugar. The glycogen content in the rest
of the body, reckoned in percentage, is increased under the action
of insulin, provided that the doses are sufficiently large. In the
case of smaller doses — except possibly in the experiments under

150

group 1 — , there is no certain change. As regards the effect oj
insulm on the total amount oj glycogen per unit of body weight,
there is some difference under different experimental eonditions.
In tlie animals which receive an injeetion of insulin without
previous fasting a decrease of the total amount of glycogen is
found, in the others no change.

Chapter 18 contains a report on snpplcmmiiary experiments on
rats. The results are in complete accordance with the thorough
investigations previously made bj' CoRl and CoRi. These animals
too — if any noticeable changes are found as the result of the
insulin injection — , show a fall in the liver glycogen content and
a rise in the glycogen content of the muscles and skin. The fall
in the muscle glycogen content observed in fasting animals after
large doses of insulin seems to be due to the hypoglycemic con-
vulsions.

Chapter H gives an account of experiments on rabbits. The
reports in the literature in regard to the effect of insulin on rab-
bits are very confusing. As regards the liver, some authors report
an increase of the glycogen content under the action of insulin,
others a decrease. This seems to be due to the fact that the liver
glycogen content in the rabbit, even under standardized experi-
mental conditions, is subject to very marked individual varia-
tions. This, in conjunction w'ith the fact that most investigators
have examined a very small number of rtibbits, seems to be the
explanation of the divergent results.

The author’s investigations, Avhich were especially intended to
control the statements to the effect that in young starving rabbits
insulin brings about a storage of glycogen in the liver, indicate
that this is not the case. Rabbits, like other experimental animals,
react to a supply of insulin — if it produces any effect at all — .
by a jail of the liver glycogen content. This result, of course, is
exactly the reverse of that found e.g by Goldblatt and
CoRKiLL, who nevertheless worked with rather large series. The
reason for this divergence is not clear, but their results may be
accounted for by some overlooked source of error. It would be
very peculiar if one kind of animal should react to insulin in an
essentially different way from others.

151

Bibliography

Abderhalden, E. and E. Wertheimer, Arch. f. d. ges. Physiol. 1932, 280i 601.

Arch. f. d. ges. Physiol. 1934, 235, 53.

Abelin, L, Ztschr. Vitaminforch. 1935, 4, 120.

Adamoff, V., Ztschr. f. Biol. 1904 — 05, JfG, 281.

Ahlgren, G., Arch. f. d. ges. Physiol. 1925, 47.

Aldehof, G., Ztschr. f. Biol. 1888—89, 25, 137.

Alienburger, E., Klin. Wchnschr. 1936, 15, 1129.

Anderson, 1. A. and J. J. R. Macleod, Biochem. J. 1930, £4, 1408.
Arndt, H.-J., Beitr. path. Anat. u. allg. Path. 1927, 79, 69.

Babkin, B. P., Brit. J. Exp. Pathol. 1923, 4, 310.

Barbour, A. D., I. L. Chaikof, J. J. R. Macleod and M. D. Orr, Am. J.
Physiol. 1927, 80, 243.

Barfnrth, D., Arch. mikr. Anat. 1885, 25, 259.

Barrenschcen, H. K., Biochem. Ztschr. 1913, 58, 277.

Bauercisen, E., Ztschr. f. d. ges. exper. Med. 1938, 103, 145.

Bernard, C., Cited from Pfliiger (1905).

Bernhard, F., Biochem. Ztschr. 1925, 157, 396.

Best, C. H., J. P. Hoet and H. P. Marks, Proc. Boy. Soc. London Ser. B.
1926, 100, 32.

— and N. B. Taylor, »The Physiological Basis of Medical Practice®, 1943.
Bickel, A. and J. A. Collazo, Dtsche med. Wchnschr. 1923, 49, 1408.
Blaschko, H., Handb. d. Biochem. Erg.-W. 1933, 1, 2, 932.

Blatherwick, N. R., P. J. Bradshaw, 0. S. Cullimore, M. E. Ewing, H. TF.

Larson, J. Biol. Chem. 1936, 113, 405,

Babbit, B. G. and H. J. Deuel Jr., Am. J. Physiol. 1940, 131, 521.

Bodo, R. and H. P. Marks, J. Physiol. 1928, 65, 48.

Boehm, R., Arch. f. d. ges. Physiol. 1880, 23, 44.

— and Eojjman, Arch. f. d. ges. Physiol. 1880, 23, 205.

Bokelmann, 0. and H. Dieckmann, Klin. Wchnschr. 1930, 9, 1211.

, C, Kaujmann and Scheringer, Arch. Gyn. 1931, 146, 167.

Bomskov, C. and K. N. von KauUa, Ztschr. f. d. ges. exper. Med 1942,

no. 603.

Bonn, F. L., Med. Klin. 1925, 21, 327.

Bonnier, G. ock 0. Tedin, sBiologisk variationsanalys®. Stockholm 1940.
Borell, U. and Hj. Holmgren, to be published in Acta med. Scandinav.
1945.

Brentano, C., Klin. Wchnschr. 1939, 18, 42.

Bridge, E. M., Bull. Johns Hopkins Hosp. 1938, 62, 408.

152

Brugsch, Th., A. Bcnatt, H. Horsters and R. Katz, Biochem. Ztschr. 1924,
U7. 117.

Bum, J. II. and II. W. Ling, Quart. J. Pharmacy and Pharmacol. 1929,

2 , 1 .

Burger, M. and II. Kohl, Arch. exp. Path. 1935, 178, 269.

Choi, Y. 0., Am. J. Physiol. 1928, S3, 406.

Christo}, P., L. Iledon, A. Loubatieres and P. Monnier, Ann. de physiol.
1937, 13, 997.

Compt. rend. Soc. de biol. 1938, 127, 33.

Compt. rend. Soc. de biol. 1938, 127, 581.

Chrometzlca, F. and G. Bcutmann, Klin. Wchnschr. 1940, 19, 196.

Collazo, J. A., M. Handel and P. Rubino, Dtsche med. Wchnschr. 1924,
50, 747.

Corey, E. L. and S. TP. Britton, Am. J. Physiol. 1941, 131, 783.

Cori, C. F. and G. T. Con, Proc. Soc. Exper. Biol. Med. 1922 — 23, 20, 409.
and G. TT^ Piicher, J. Pharmacol, and Exper. Therap. 1923, 21, 377.

— J. Pharmacol, and Exper. Therap. 1925, 25, 1.

— and G. T. Cori, J. Biol. Chem. 1920, 70, 557.

— J. Biol. Chem. 1926, 70, 577.

— Harvey Lect. 1928, 23, 76.

— and G. T. Cori, J, Biol. Chem. 1928, 79, 321.

J. Biol. Chem. 1928, 76, 755.

Biochem. Ztschr. 1929, 206, 39.

J. Biol. Chem. 1929, 85, 275.

J. Biol. Chem. 1930, 100, 323.

— Physiol. Rev. 1931, 11, 143.

Cori, G. T., Am. J. Physiol. 1930, 94, 557.

— , C. F. Cori and K. IP. Buchwald, 3. Biol. Chem. 1930, 86, 375.

Coricill, B., Biochem. J. 1930, 24, 779.

Corkill, A. B., H. P. Maries and TT^ E. White, J. Physiol. 1933, 80, 193.
Crabtree, G. and B. B. Longwcll, Proc. Soc. Exper. Biol, and Med. 1936,
34, 705.

Cramer, A., Ztschr. f. Biol. 1888, 24, 67.

Croftan, A. C., Arch. f. d. ges. Physiol. 1908, 126, 407.

Daoud, K. M. and H. A. F. Gohar, J. Physiol. 1933, 80, 314.

Davenport, H. A. and H. K. Davenport, 3. Biol. Chem. 1927, 76, 651.
Deuel, H. J. Jr., M. Gulick, C. F. Gntnewald and C. H. Cutler, 3. Biol.
Chem. 1934, 104, 519.

Deuel, H. J., L. F. Hallman and S. Murray, 3. Biol. Chem. 1937, 119, 257.
Deuel, H. J. Jr., L. F. Hallman, S. Alurray and L. T. Samuels, 3. Biol.
Chem. 1937, 119, 607.

— , J. S. Butts, L. F. Hallman, S. Murray and H. Blunden, 3. Biol. Chem.
1937, 119, 617.

J. Biol. Chem. 1938, 123, 257.

Dittmar, Arch. f. exper. Pathol, und Pharmakol. 1933, 171, 496.

153

Donaldson, H. H., J. Comp. Neurol, and Psychol. 1908, 18, 346.

Dudley, H. Tl^. and G. F. Biochem. J. 1923, 17, 435.

Edlund, F. and Hj. Holmgren, Ztschr. f. d. ges. exper. Med. 1941, 109, 11.

— Ztschr. f. d. ges. exper. Med. 1942, 111, 281.

Eger, 11^. and V. Morgenstem, Arch. f. e.vper. Path, und PhTmakoI 1938,
191, 76.

— Virchows Arch. f. path. Anat. 1942, 309, 607.

Ekrismann, 0., Arch. f. exper. Path. u. Pharmakol. 1927, 121, 299.

Eislcr, B. and R. Hemprich, Ztschr. f. d. ges. exper. Med. 1932, 83, 439.
Elias, H. and E. Schvbert, Biochem. Ztschr. 1918, 90, 229.

Epstein, A. A. and P. IF. Aschner, J. Biol. Chem. 1916, 25, 151,

— , J. Reis and J. Branower, J. Biol. Chem, 1916, 2G, 25.

V. Euler, U. S. and A. G. Holmquist, Arch. f. d. ges. Phj'siol. 1934, 231),

210 .

— sStudien uber die Beeinflussung der Gewebeoxydation durch Adrenalin
und Insulins. Inaug. Diss. Stockholm 1930.

Evans, C. L., C, Tsai and F. G. Young, J. Physiol. 1931, 73, 67.

Fichera, G., Med. sper. Arch. ital. 1940, 6, 49.

Fiessinger, N., H. Benard, M. Herbain, L. Dernier and G. Barcillier, Compt.

rend. Soc. de biol. 1937, Hi, 952.

Ann. de med. 1939, 45, 5.

— Rev. stiint. med. 1940, 29, 411.

Fisher, N. F. and R. W. Lackey, Am. J. Physiol. 1925, 72, 43.

Fisher, R. A., sStatistical Methods for Research Workers*, Ed.: 6, Lon-
don 1936.

Folin, 0., H. C. Trimble and L. H. Newman, J. Biol. Chem. 1927, 75, 263.
Fomaroli, P. and 1. Boni, Boll. soc. ital. biol. sper. 1940, 15, 515.
Forsgren, E., »Mikroskopiska och experimentella leverundersbkningar*,
Inaug. Diss., Stockholm, 1927.

— Skand. Arch. f. Physiol. 1929, 55, 144.

Frank, E., E. Hartmann and M. Nothmann, Klin. Wchnschr. 1925, i,
1067.

Arch. {. exper. Path. u. Pharmakol. 1927, 127, 35.

Fitjit. /., Tohoku J. Exper. Med. 1924, 5, 405.

Funk, C., Ztschr. f, physiol. Chem. 1914, 89, 373.

Gain, T. and L. Raffo, Arch, di patol. e. din. med. 1939, 20, 109.

Gamier, L. and M. Lambert, Compt. rend. Soc. de biol. 1897, i, 718.
Geiling, E. M. K., H. Jensen and G. E, Farras, Handb. exp. Pharmakol.
1937, Ergv'. 5, 197.

Gemmill, C. L., Bull. Johns Hopkins Hosp. 1940, 66, 232.

— and L. Hamman Jr., Bull. Johns Hopkins Hosp. 1941, 68, 50.
Gerritzen, F., Nederl. Tijdschr. Geneesk. 1942, 86, 1599.

Gierke, E., Verhandl. d. deutsch. path. Gesellsch. 1907, 10, 182.

Goldblatt, M. W., Biochem. J. 1927, 21, 991.

— Biochem. J. 1929, 23, 83.

154

— Biochem. J. 1929, 23, 243.

— Biochem. J. 1930, 24, 1199.

— J. Physiol. 1933, 79, 280.

Graymail, /., Am. J. Physiol. 1941, 133, 300.

Greisheimer, E. M. and 0. II. Johnson, Am. J. Physiol. 1929, 90, 369.

Am. J. Phj'siol. 1930, 93, 653.

Am. J. Physiol. 1930, 94, 11.

Grevenstvk, A. and E. Laqueur, Biochem. Ztschr. 1925 a, 163, 390.

Ergebn. d. Physiol. 1925 b, 23, 1.

Biochem. Ztschr; 1926, 173, 283.

Grube, K., Arch. f. d. ges. Ph}'siol. 1905, 107, 483.

— Arch. f. d. ges. Physiol. 1907, 118, 1.

Guest, M. M. and R. A. Rawson, Am. J. Physiol. 1939, 126, P 515.
Gvlick, M., L. T. Samuels and H. J. Deuel Jr., J. Biol. Chem. 1934, 105, 29.
Giirber, A., Sitzb. phys.-med. Gesellsch. Wiirzbg. 1895, 17.

Handowsky, H., Arch. f. expsr. Path. u. Pharmakol. 1928, 134, 324.

— and K. Wcst-phal, Arch. f. d. ges. Physiol. 1928, 220, 399.

Ilax, D., Arch. f. d. ges. Physiol. 1927, 216, 627.

11 ebb, C. 0., Quart. J. Exper. Physiol. 1938, 27, 237.

Ilechter, 0., R. Levine and S. Soskin, Proc. Soc. Exper. Biol. Med. 1941,
41i, 390.

Hcnschen, C., Arch. f. klin. Chir. 1932, 173, 488.

Hermann, V. S., Ztschr. f. physiol. Chem. 1938, 251, 78.

— Ztschr. f. physiol. Chem. 1939, 262, 95.

Heymann, TF. and J. L. Modic, J. Biol. Chem. 1939, 131, 297.

Heymans, B. and C, Heymans, Compt. rend. Soc. de biol. 1925, 93, 50.
Hines, H, M., C, E. Leese and A. P. Barer, Proc. Soc. Exper. Biol. Med.
1928, 25. 736.

Hirsch, G. C. and R. F. J. van Pelt, Proc. K. Akad. Wetensch. Amsterd.
1937, .lO: 2, 538.

Hoffman, A. and E. Wertheimer, Arch. f. d. ges. Physiol. 1927, 217, 728.
Holmgren, Hj., Acta med. Scandinav, Suppl. 74, 1936.

— Ztschr. f. d. ges. exper. Med. 1941, 109, 315.

— Acta Zool. 1944, 25, 169.

Holmgren, Hj. and S. Wohlfahrt, Acta Psych, et Neurol. Suppl. 31, 1944.
Holmquist, A. G., Ztschr. f. mikr.-anat. Forsch. 1931, 25, 30.

Hrubetz, M. C. and S. N. Blackberg, Am. J. Physiol. 1938, 122, 759.
Isaac, S. and R. Siegel, Hand. d. physiol. 1928, 5, 469.

Ishimori, K., Biochem. Ztschr. 1913, 48, 332.

Jonen, P., Ztschr. Kinderheilk. 1924, 38, 46.

Junkersdorf, P., Arch. f. d. ges. Physiol. 1921, 186, 254.

Kapfhammer, J., Handb. d. Biochem. 1927, 9, 98.

Karczag, L., J. J. R. Macleod and M. D. Otr, Proc. Roy. Soc. Canad.
1925, 19, Sect. V, 57.

Kaunitz, H. and L. Selzer, Ztschr. f. d. ges. exper. Med. 1937, 101, 111.
Kepinov, L., Compt. rend. Soc. de biol. 1938, 127, 851.

155

McKay, M., Proc. Roy. Soc. Canad. Sect 5, 1928, 22, 329.

Kerly, M., Biochem. J. 1930, 2^., 67.

Kissel, 0., Verhandl. d. phys-med. Gesellsch. zu Wiirzbg. 1896, 30, 77.
Kobori, B., Biochem. Ztschr. 1926, 173, 54.

Kiilz, E., Arch. f. d. ges. Physiol. 1881, 2^, 46.

— Arch. f. d. ges. Physiol. 1881, 24, 57.

Kiilz, R., Ztschr. f. Biol. 1886, 22, 161.

Kunz, H. and H. Molitor, Arch. f. d. ges exper. Path. 1929, 145, 211.
Lajos, S., Biochem. Ztschr. 1936, 284, 279.

Ldnczos, A., Arch. f. d. ges. Physiol. 1934, 233, 787.

— Arch. f. d. ges. Physiol. 1935, 235, 422.

Laqueur, E. and S. E. de Jcmgh, Biochem. Ztschr. 1925, 163, 308.
Lattber, H. J., Arch. f. klin. Chir. 1938, 193, 74.9.

— and T. Bersin, Klin. Wchnschr. 1939, 18, 232.

Laujberger, V., Ztschr, f. d. ges. exper. Med. 1924, 42, 570.

Lawrence, R. D. and R. A. McCane, Biochem. J. 1931, 25, 570.

V. Ledebur, J., Handb. d. Biochem. 1936, Ergw. 3, 907.

Liebig, H., Arch. f. exper. Path, und Pharmakol. 1940, 195, 617.

Livini, F., Compt. rend. I’anoc. d. anat. 1927, 22, 144.

Lockkead, J. and B^. Cramer, Proc. Physiol. Soc. 1906, 35, 11.

Loeh, R. F., E. G. Nichols and- B. H. Paige, Arch. Int. Med. 1931, 48, 70.
Long, C. N. H., J. Biol. Chem. 1928, 77, 563.

— and G. T. Evans, Proc. Soc. Exper. Biol. Med. 1932, 30, 186.
Lubarsch, 0., Virchows Arch. f. path. Anat. 1906, 183, 188.

Lnchsinger, B., sExperimentelle und kritische Beitrage zur Physiologic und

Pathologic des Glykogensu. Inaug. Diss. Zurich 1875.

Lundsgaard, E., XJpsala Lakaref. forh. 1939, 4-5, 143.

Low, A. and A. Krcrna, Biochem. Ztschr. 1929, 206, 360.

Macleod, J. J. R. and R. G, Pearce, Am. J. Physiol. 1911, 27, 341.

— Physiol. Rev. 1924, 4> 21.

Mann, F. C., Harvey Lectures 1927, 23, 49.

— and T. B. Magath, Ergebn. d. Physiol. 1924, 23, 212.

Marchand, F., Arch. f. path. Anat. 1885, 100, 42.

Markowitz, J., Am. J. Physiol. 1925, 74, 22.

Masayama, T. and 0. Riesscr, Biochem. Ztschr. 1931, 234, 323.

May, F., Ztschr. f. Biol. 1934 a, 95, 277.

— Ztschr. f. Biol. 1934 b, 95, 606.

— and H. Weinbrenner, Ztschr. f. Biol. 1938, 99, 199.

Meixner, K., Beitr. zu gerichtl. Med. 1911, 1, 222.

Merten, R., Ztschr. {. d. ges. exper. Med. 1939, 105, 273.

V. Meyenhurg, H., Schweiz, med. Wchnschr. 1924, 54, 1121.
de Meyer, J., Arch, intemat. physiol. 1909, 8, 204.

Mirski, A., I. Rosenbaum, L. Stein and E. Wertheimer, J. Physiol. 1938,
92, 48.

Molitor, H. and M. Poliak, Arch. f. exper. Path, und Pharmakol. 1930,
154. 280.

156

MorelU, A. and L. d’ Ambrosia, Arch, di sc. biol. 1938, £4, 331.

Moscati, G., Beitr. zu Chem. Physiol, u. Pathol. 1907, 10, 337.

Nasse, 0., Arch. f. d. ges. Physiol. 1877, 14, 473.

Neubauer, 0., Handb. d. biol. Arbeitsmethoden 1925, Abt. 4, Teil 9, 583.
Neujcld, A. H. and J. B. Collip, Endocrinology 1941, 28, 926.

— , S. M. Scoggan and G. S. Stewart, Endocrinology 1940, 27, 132.
l^utter, P. E., J. Nutrition 1941, 21, 477.

Orten, J. M. and G. Sayers, J. Biol. Chem. 1942, 145, 123.

Osborne, T. B. and L. B. Mendel, J. Biol. Chem. 1924, 59, 13.

Palladin, A., Biochem. Ztschr. 1924, 152, 228.

Paulesco, N. C., Compt. rend. soc. de biol. 1913, 74, 627.

Petren, T., Morphol. Jahrb. 1939, 83, 256.

Pfliiger, E., Arch. f. d. ges. Physiol. 1902, 91, 119.

— Arch. f. d. ges. Physiol. 1903, 95, 19.

— Arch. f. d. ges. Physiol. 1907, 119, 117.

— Arch. f. d. ges. Physiol. 1909, 91, 119.

— »Das Glykogen*. 2 Aufl. Bonn 1905.

Pincussen, L., Klin. Wclinschr. 1932, 2, 1231.

Rafferty, M. A. and P. L. MacLachlan, J. Biol. Chem. 1941, 140, 167.
Rathery, F., S. Gibert and Y. Laurent, Compt. rend. soc. de biol. 1930,
104, 652.

Compt. rend. soc. de biol. 1931, 107, 1264.

Compt. rend. soc. de biol. 1931, 107, 1490.

Reid, C., J. Physiol. 1936, 87, 121.

Richter, F., Beitr. zu path. Anat. u. zu allg. Path. 1931, 86, 65.

Rossi, G., Ber. ges. Physiol, u. Pharm. 1928, 43, 65.

Russell, J. A., Am. J. Physiol. 1930, 124, 774.

Sahyun, M. and J. M. Luck, J. Biol. Chem. 1929, 85, 1.

— , R. Simmonds and H. WorMng, Am. J. Physiol. 1934, 108, 708.
Salomon, G., Centralblatt med. Wissensch. 1874, 12, 738.

Sato, K., Tohoku J. Exper. Med. 1923, 4> 265.

Scheiff, W., Arch. f. d. ges. Physiol. 1931, 226, 481.

Schenk, P., Arch. f. exper. Path, und Pharmakol. 1923, 99, 206.
Schondorff, B., Arch. f. d. ges. Physiol. 1903, 99, 191.

Scoz, G., Boll. soc. ital. di biol. sperim. 1929, 4> 1158.

Seckel H. P. G., Endocrinology 1938, 23, 760.

— and K. Kato, Arch. Pathol. 1938, 25, 347.

Seegen, J. and F. Kratschmer, Arch. f. d. ges. Physiol. 1880, 22, 223.
Shimamura, M., Jap. J. Med. Sc. IV Pharm. 1938, 12, 22.

Silvette, H. and S. W. Britton, Am. J. Physiol. 1932, 100, 685.

Simpson, W. W. and J. J. R. Macleod, J. Physiol. 1927, 64, 255.

Sjogren, B., T. Nordenskold, lij. Holmgren and J. Mollerstrom, Arch,
f. d. ges. Physiol. 1938, 240, 427.

Snedecor, G. W., »Statistical Methods*. Third Ed. Ames, Iowa 1940.
Spitzbarth, H., Wien. klin. Wchnschr. 1940, 53, 1031.

157

Stadie, TT^ C., F. D. W. Lukens and J. A. Zajyp Jr., J. Biol. Chem. 1939,
128, 98.

Staub, H., Handb. norm. Physiol. 1930, 16: 1, 557.

Stein, L., E. TuerJdscher and E. Wertheimer, J. Physiol. 1939, 95, 350.
Steinmetzer, K. and F. Siooboda, Biochem. Ztschr. 1928, 198, 259.

Stieve, H. and ZI. Kaps, Ztschr. f. mikr-anat. Porsch. 1937, 42, 499.
rStudentr) (W. S. Gosset), Biomelrica 1908,6, 1.

Sunaba, ¥., Mitt. a. d. med. Akad. zu Kioto 1936, 17, 336.

Sandberg, C., Ztschr. {. d. ges. Anat. 1924, 73, 168.

Tacki H. and K. Takai, Tohoku J. Exper. Med. 1926, 7, 411.

Tcherkes, L. A. and E. L. Rosenjeld, Biochimija 1941, 6, 58. Kgrzbl. 1941,
109, 234.

Terada, K., Tohoku J. Exper. Med. 1939, 36, 180.

Terbriiggen, A., Virchows Arch. f. path. Anat. 1937, 298, 646.

Tonutti, E. and J. Wallrafj, Ztschr. f. mikr.-anat. Forsch. 1938, 44, 532.
Vendeg, V., Arch. f. d. ges. Physiol. 1935, 235, 674.

Fjsco, S., Atti accad. Lincei 1926, 4, 153- Endocrinology 1926, 11, 79.
van Weel, P. B., Genecsk. Tijdschr. Nederl.-Indie 1941, 81, 2297. Cite
from Deuel, H. J. Jr., Ann. Rev. Biochem. 1943, 12, 135.

Wertheimer, E., Arch. f. d. ges. Physiol. 1928, 219, 190.

Wesselkina, M., Ztschr. f. d. ges. exper. Med. 1932, 85, 463.

Wetzel, R., H. Wqlhckitt, H. Rvska and Tk. Oestreicher, Arch. f. exper.

Path. u. Pharmakol. 1936, 181, 703.

V. Wittich, Centralbl. med. Wissensch. 1875, 13, 113.

Agren, G., 0. Wilander and E. Jorpes, Biochem. J. 1931, 25; 777.

ACTA PHYSIOLOGICA SCANDINAVICA
VOL. 11 SUPPLEMENTVM XXXIV

FROM THE DEPARTMENT OF PHYSIOLOGY, UNIVERSITY OF LUND, SWEDEN

ON THE PRESENCE OF HISTAMINE IN
PLASMA IN A PHYSIOLOGICALLY
ACTIVE FORM

NILS EMMELIN

LUND

19 4 5

Contents

Chapter 1. Introduction 5

Chapter 2. Method of quantitative estimation of histamine in blood

plasma 10

Chapter 3. Identification of the g^t contracting substance in ultra-
filtrates of plasma IG

Chapter 4. Is all the blood histamine contained in the corpuscles? . . 25

Chapter 5. Is the plasma histamine present in a physiologically

active or inactive form? 35

Chapter 6. Experiments where the histamine content of plasma is

artificially increased 50

Chapter 7. Discussion 58

Summary' 64

Acknowledgement 66

References 67

Chapter 1

Introduction

Only a few years after the synthesis of /9-imidazolylethyIaniine
it was discovered that this substance has remarkable biological
effects, contracting plain muscles, dilating capillaries and sti-
mulating glandular activity (Dale and Laidlaw 1910, Popielski
1920). About the same time it was shown that this active base
normally occurs in the body and seems to be an ordinary
constituent of the cell (Barger and Dale 1911, Abel and Kubota
1919, Best, Dale, Dudley and Thorpe 1927). Resulting from
these discoveries it was suggested that histamine was engaged
in several normal and pathological processes in the body. Thus
histamine is thought to play a part in the local regulation of
the blood flow through the tissues, e. g. causing vasodilatation
in reactive hj'-peraemia and in muscular activity. Further hista-
mine has been referred to as »the hormone for gastric secretions
and has also been considered a normal stimulant of intestinal
motility. Examples of pathological conditions where histamine
has been thought to play a more or less important role are:
surgical shock, shock secondary to burns, peptic ulcer, toxaemias
in pregnancy, anaphylactic shock, hypersensitiveness to cold,
urticaria, hay fever, Meniere’s syndrome, a special type of head-
ache, intestinal autointoxication etc. Histamine is also supposed
to be responsible for some of the symtoms produced by snake
venoms, bee venom and the toxins of certain bacteria. A great
number of monographs and articles dealing with theories con-
cerning the part played by histamine in these and other reactions
have been published (Dale 1920, 1929, 1933, Feldberg and Schilf
1930, Kiipper 1930, Best and McHenry 1931, Gaddum 1936,
Guggenheim 1940, Rigler 1938, Code and MacDonald 1937,

Fel'dberg 1937, Gajdos 1938, Marcou et al. 1938, Ahlmark and
Kornerup 1939, Dragstedt 1940, 1945, Rocha e Silva 1944).

These theories aroused interest in the estimation of histamine
occurring in blood and especially in blood plasma. Although
it is believed that under physiological conditions histamine acts
locally as »Grewebshormon», many investigators have tried to
trace the histamine instrumental in these reactions in the venous
blood emerging from the tissue. Already at an early stage of
investigation in this field it was believed that in pathological
processes, especially in severely damaged tissue, histamine is
liberated locally, absorbed into the blood stream, and carried
to distant organs causing specific reactions in the periphery.

These investigations led to the discovery that histamine is a
normal constituent of blood. The experiments of Harris (1927)
and Best and McHenry (1930) suggested that histamine is
normally present in the blood. This was, however, denied by
Koessler and Hardee (1924),, Guttentag (1931), Zipf (1931), Zipf
and Hiilsmeyer (1933), Mac Gregor and Peat (1933) and Mac
Gregor and Thorpe (1933). After Barsoum and Gaddum (1935)
had devised a method for the estimation of histamine — later
modified by Code (1937) — the presence of histamine in normal
blood must be considered to be well founded. In their first paper
on this subject Barsoum and Gaddum also demonstrated the
presence of histamine in plasma by this method, a finding which
has been confirmed by later investigators using similar methods.

The fact that histamine is present in plasma is rather puzzling
from several points of view. It is remarkable that this highly
active substance can be found in plasma in relatively high
concentrations. In the guinea-pig the plasma contains approxi-
mately 200 y of histamine per litre plasma or even more. If a
saline solution containing histamine in this concentration is per-
fused through the blood vessels or through the bronchial tree of an
isolated guinea-pig’s lung it elicits a marked bronchoconstriction
(Bartosch, Feldberg and Nagel 1932, Epstein 1932). A piece of
guinea-pig’s ileum, suspended in Tyrode’s solution, usually con-
tracts strongly when histamine is added in a quantity giving
a concentration of 5 y per litre in the bath; if the concentration
is increased to e. g. one fourth of the plasma concentration a

maximal contraction is likely to occur. It is also striking that
minute amounts of histamine, when injected intravenously, are
capable of causing large physiological effects although the
plasma of the animal contains histamine in a relatively high
concentration. In fact, on injection of amounts causing large
physiological effects, the increase in the histamine content of
the blood or plasma is so small that it can not be detected
by the methods available (Rose 1940, Emmelin, Kahlson and
Wicksell 1941).

It is also remarkable that histamine occurs in the plasma in
spite of the presence of a histaminolytic factor in the blood.
In several species where the plasma has a histaminolytic activity,
the plasma gives histamine on extraction; in various instances
we have found histamine in the plasma of pregnant women who
are at such a period when the histaminolytic power of plasma
is known to be verj'’ great (for a detailed report on the hista-
minolytic power of plasma see Ahlmark 1944). Not only plasma
but also certain tissues contain agents which should make it
impossible for histamine to be present in plasma. Best and
McHenry (1930) found that a dog’s kidney perfused with blood
containing added histamine was, within four hours, able to
destroy as much as 200 mg of histamine. A heart-lung-kidney-
preparation is capable of inactivating 25 mg of histamine in 15
minutes (Steggerda, Essex and Mann 1935). A high histamino-
lytic power in isolated perfused organs has also been found by
Mac Gregor and Peat (1933) and Sibul (1935). The well-known
fact that the effect of intravenously injected histamine is very
transient is partly due to the histamine inactivating power of
the organism: the histamine content of the plasma rapidly
decreases to its original level.

It is thus obvious that the alleged appearance of histamine
in blood plasma gives rise to a series of problems. Some of the
ways in which these problems may be dealt with will be con-
sidered here.

The first question arising is this: is it proved beyond doubt
that the substance found in 'plasma is really histamine? It must
be born in mind that the biological methods used for detection
of histamine are not specific. The presence of histamine in

blood seems to be well established: estimations with different
test objects and certain specific reactions all point in this
direction; besides, using rabbit’s blood, which is especially rich
in histamine. Code and Ing (1937) were able to isolate and
chemically identify histamine from »the white layer». This is,
however, not directly applicable to the small part of the blood
histamine which is claimed to be present in plasma; so far, it
can only be stated that plasma, treated according to the usual
methods of extraction, stimulates the isolated guinea-pig’s gut
and the fowl’s rectal coecum. Although is seems probable that
the substance in question is histamine, further attempts to
identify the agent appear necessary.

A second method of treating the problems is suggested by the
following facts. Barsoum and Gaddum (1935), investigating
rabbit’s blood, found about 15 per cent of the blood histamine
in plasma. Anrep and Barsoum (1935) and Code (1937), con-
tinuing this work, found a somewhat smaller part of the blood
histamine in plasma, and Minard (1941) stated, that only 2 — 3
per cent of the histamine is contained in the plasma. It is
obvious that the more the methods have been improved, the
smaller has become the part of the total blood histamine which
can be found in plasma. It has been demonstrated that most
of the blood histamine, at least in the rabbit, is present in the
platelets (Schwartz 1936, Zon, Ceder and Crigler 1939, Minard
1941). Considering the fragility of the platelets it is reasonable
to suspect that physiologically plasma does not contain any
histamine at all; the small quantity of histamine found in a
plasma sample might have been liberated at the breakdo^vn of
some few platelets during the collection and centrifuging of
the blood samples; even the slightiest tendency of clotting will
have the same effect. Particularly where the rabbit is concerned
this interpretation seems reasonable as the platelets of this
animal are exceptionally rich in histamine. Minard (1941)
suggests such a possibility. With reference to human blood
this possibility has, under somewhat different aspects, been
discussed by Barsoum and Smirk (1936).

A third possibility is, that histamine is really contained in
the plasma but is present there in such a form that it is

physiologically inactive and protected against the histamino-
lytic effect of blood and tissues. It should be emphasized that
the biological estimation of the histamine content of plasma is
proceeded by drastic chemical extraction processes. Gaddum
(1936) points out that the demonstration of histamine in blood
in this way »would not justify the view that histamine is present
in a free and active form . . . These extracts are subjected to
prolonged boiling in the presence of strong acid. . . . one of its
effects may be to liberate histamine from an inactive precursor.
The evidence on this latter point is still incomplete.» Also Dale
(1937, 1938) and Feldberg (1937) have emphasized that the
available methods do not answer the question whether histamine
is present in blood in a free and active form or in a combined,
inactive state.

The following investigation is an attempt to treat these pro-
blems from the three points of view which have been discussed
above.

Chapter 2

Method of quantitative estimation of histamine
in blood plasma

In these experiments we have used plasma from guinea-pigs,
rats, rabbits, cats and dogs. The chemical methods available
are obviously not sensitive enough for the detection of the small
amounts of histamine in plasma. In this chapter we will only refer
to the guinea-pig gut method, which was used as a routine. Other
test objects have also been used, but as this has been done chiefly
to identify the histamine they will be discussed in chapter 3.
Tlie method for estimation of histamine in plasma ndth the
guinea-pig’s ileum as a test object, such as is practised in this
laboratory, has been described previously (Emmelin, Kahlson
and Wicksell 1941) and a few details only will be added here.

Assay on isolated guinea-pig’s gut, Zadina (1939) stated that
the gut should be taken from young guinea-pigs, weighing 200
— 250 g. Like Ahlmark (1944) we have come to the conclusion,
that it is preferable to use larger animals; the gut of animals
weighing more than 300 g is more sensitive to histamine and
it also gives constant responses to the same dose of histamine
for a much longer period. It is often claimed that the animal
should be starved for some time before the assay; our opinion
is that this is not of any importance.

In several experiments we have worked on preserved gut.
Guggenheim and Loffler (1916), who introduced the guinea-pig
gut as test object for histamine, suggested such a method, later
recommended by Minard (1941).

After the guinea-pig has been killed by a blow on the head an
empty piece of gut is removed. The gut is kept in Tyrode solution in
the refrigerator, — 2 hours before the assay a small piece is put

into warm Tyrode solution, through which oxygen is bubbled. The gut

soon begins to contract spontaneously and can then be used in the
ordinary way. Pieces of the same gut, presented in this manner, may
often be used 4 — 5 days after the animal has been killed.

The gut is suspended in a bath containing about 2 ml of Tyrode
solution. Oxygen holding 1 per cent of carbon dioxide is blown through
the bath.

The Tyrode solution is prepared in the following way. To 40 ml of
solution I and 20 ml of solution II aq. dost, ad 1000 ml is added.
Solution I: NaCl 200 g, KCl 5 g, GaCl- + 6HjO 10 g, JlgCU + 6H=0 5 g,
aq. dest. ad 1000 ml. Solution II; NallsPO* 1.25 g, aq. dest. ad 500 ml.
This solution also contains NaHCOs in a quantity, which is determined
each time new base solutions are prepared; this quantity is chosen so
as to give the Tyrode solution when thoroughly bubbled with oxygen
holding 1 per cent COs, a pll of 7.i — ^7.3 (determined electrometrically).

Kwiatkowski (1941), testing histamine on guinea-pig’s gut with a
special perfusion method, described by Gaddum, Jang and Kwiatkowski
(1939), suggested a Tyrode solution with only one half of the usual
amount of CaCli. In this way the sensitivity of the gut to histamine
was greatly increased. McDowall and MeWhan (1936) have also noticed
that histamine effects are more pronounced in a solution poor in calcium.
We have tried this solution in the usual test system but abandoned it
since we have found that the assay is rendered more difficult by the
impaired relaxation of the gut. Addition of glucose to the Tyrode
solution, also recommended by Kwiatkowski, increased neither the sen-
sitivity nor viability of the gut in our experiments.

In some experiments the Tyrode solution has contained atropine
sulphate in the concentration of 1 : 2 millions, but we are not convinced
that this is of any advantage when examining blood plasma; surely it
greatly reduces the sensitivity of the gut to histamine.

In the biological assay the plasma samples have been matched
against a standard solution of histamine biphosphate, containing
100 Y of histamine base per litre. This solution has been freshly
prepared before each experiment. All values in this paper are
given in terms of histamine base.

Preparation of the plasrna for biological assay. Before the
estimation on the gut takes place the plasma samples must be
treated by a method, which leaves as much as possible of the
histamine intact but eliminates substances which interfere with
the biological assay. Two methods have been used. Firstly the
plasma has been treated according to the chemical method of
Barsoum and Gaddum (1935) as modified by Code (1937).
Code’s method has the advantage of being simpler and of -not

causing so great a loss in histamine as the original one (Code
1937, Code, Evans and Gregory 1938). Anrep et al. (1939)
claim that the Code method gives higher values when used on
blood than the Barsoum-Gaddum method because it does not
eliminate a gut contracting substance which is not histamine,
but according to Kwiatkowski (1941) potassium. In experiments
on plasma this possible error is in any case of no importance
as the interfering agent is contained in the red corpuscles
(Anrep et al. 1939, KAviatkowski 1941). After extraction the
plasma samples have been dissolved in distilled water and
neutralized with NaOH, using bromothymol blue as an indicator.
Care has been taken to obtain a pH corresponding as closely
as possible to that of the Tyrode solution in the test bath.

Apart from the chemical extraction procedure an ultrafiltra-
tion method has been used. Experiments with this method have
been described earlier (Emmelin 1945). The technique of ultra-
filtration used is given by Rehberg (1943).

The sample is contained in a cellophane tube surrounded by a basket
of metal wire, Avhich is placed in a centrifuging tube. The filtration
pressure is obtained by centrifugation. Of the different cellophane tubes
recommended by Rehberg we have used the most impermeable ones.
These are according to Rehberg absolutely impermeable to protein. The
cellophane tubes are kept in water containing formaline. The formaline
must be thoroughly removed by washing as it is known that it in-
activates histamine (Kendall 1927) and makes test objects less sensitive
to histamine (Best and McHenry 1930). There are two sizes of tubes,
one taking about 15 ml, the other about G ml. Both sizes have been
used but Ave prefer the larger Avhich offers a greater filtration surface.
As we usually have to work with small amounts of plasma we have
placed a glass rod of suitable size in the middle of the tube. In this
way we have, for instance, been able to fill one of the large tubes rvith
about 5 ml of plasma using the thickest rod thus utilizing the whole
filtration surface. The samples have been centrifuged at a speed of
3.000 — 3.500 rev. per min. for a period ranging from 30 — 120 minutes.
Using the large tube and the thickest glass rod we obtain with the
centrifuge at our disposal about 1 ml of ultrafiltrate in an hour rvith
5 ml of plasma. Now and then a defect in the Avail of the tube may
occur but this is easily discovered: a turbid, plasma coloured liquid is
obtained instead of the clear colourless ultrafiltrate. In some experi-
ments chlorazol fast pink has been used to prevent clotting and a lesion
of the tube has then been still easier to detect as this dye does not pass
the intact membrane.

The ultrafiltrate of plasma is slightly alkaline in comparison to the
Tyrode solution used. In our earlier experiments the samples were
neutralized with HCl using bromothymol blue as an indicator but this
proved unnecessary as the unneutralized and the neutralized samples
elicited the same contraction of the gut.

Preliminary experiments show that histamine easily passes
the cellophane membrane. Testing 12 ultrafiltrates of the stan-
dard histamine solution (100 y/1) we have obtained the following
values: 95, 100, 90, 90, 100, 100, 100, 100, 100, 100, 100, 100 y
histamine per litre. Histamine is also obtained from a standard
solution with added egg albumen and the yield is greater than
that of the chemical extraction method (Emmelin 1945). If
histamine is present in plasma it seems thus probable that it
should be recovered almost quantitatively in the ultrafiltrate.
The question now arises whether the ultrafiltration method also
meets the second demand: to eliminate substances which could
interfere with the biological assay.

Disturbing substances can be of different kinds. They might
cause a contraction or a relaxation or they might change the
histamine sensitivity of the gut in either direction. It should
be pointed out that the typical effect of native plasma from the
species investigated injected into the bath is a contraction of
the gut. The corresponding ultrafiltrate causes a considerably
smaller contraction (Emmelin 1945; see also figure 2 in this
paper). The untreated plasma thus Seems to contain a gut
contracting agent apart from histamine. Its nature is unknown.
It must be a compound of a large molecular size as it does not
pass the cellophane membrane; the more the sample is con-
centrated by ultrafiltration the greater becomes the contraction
released by it. This substance also differs from histamine in
other respects: it causes a contraction of the gut which is not
antagonized by thymoxyethyldiethylamine or theamine (figure 2).
Further it makes the gut less sensitive to histamine, an effect
which is reversible and disappears if the gut is washed in
Tyrode solution. This effect may of course originate from some
other plasma compound than the gut contracting one.

The ultrafiltration has apparently eliminated some gut stimu-
lating agent, but there may be others which pass the membrane.

There remains the objection that the contraction elicited by the
plasma ultrafiltrate might partly or wholly be due to some
substance other than histumine. The experiments of chapter 3
indicate that histamine is responsible for the contraction.

Neither does there seem to be any relaxing agent of high
activity present in the ultrafiltrate. If the gut is made insen-
sitive to histamine by adding thymoxyethyldiethylamine, a
plasma ultrafiltrate does not cause any relaxation (figure 2).
It should be possible in this way to discover a gut relaxing
substance (Emmelin 1943). The first compound to be suspected
in this connexion is adenylic acid; it is, however, well-known
that most of it is contained in the blood cells (Buell and Perkins
1928, Barsoum and Gaddum 1935, Billings and Maegraith 1938).
It is thus not surprising that the ultrafiltration method can not
be used on laked blood; the hemolysis may have liberated,
apart from potassium, also adenylic acid and perhaps other
interfering agents. The ultrafiltrate of plasma, however, seems
to be relatively free from disturbing substances; from the preli-
minary report as well as from chapter 5 it is evident that ultra-
filtrates and chemical extracts of plasma agree fairly well in
the biological assay. It might be possible that other gut active
substances are present in the ultrafiltrate of plasma but have
no time to interfere with the histamine contraction, as this
occurs almost instantly after the injection. It is for instance
known that bile acids, which reduce the histamine sensitivity
of the gut do not interfere with the estimation of histamine if
the two substances are given simultaneously (Emmelin 1943).

In preliminary experiments we found that ultrafiltrates of
human plasma relatively often contain an agent which interferes
with the assay on gut, and human plasma has therefore not
been used in this work. It is possible that the ultrafiltration
method can also be used in experiments on plasma from human
beings if the assay is performed on a test object which is not
affected by this gut disturbing agent, e. g. perfused guinea-pig’s
bronchi; we are experimenting on this subject. Also in ultra-
filtrates of cat’s plasma an interfering substance may rarely
occur as can be seen from chapter 5.

It is obvious that in experiments where some foreign sub-

stance has been added to the blood it is necessary to make sure
whether this substance has been removed by the ultrafiltration
and if not, whether it interferes with the assay. We have given
special attention to anaesthetics and anticoagulants. In most
experiments on anaesthetized animals chloralose was used.
Control experiments show that chloralose in a moderate con-
centration does not alter the histamine sensitivity of the gut.
Amongst the anticoagulants chlorazol fast pink has the advan-
tage of not passing the ultrafiltration membrane. Heparine in
concentrations generally used does not affect the gut. In earlier
experiments in this laboratory it was found that extracts of
heparine sometimes had a histamine-like effect on the gut. We
have therefore tested each sample of heparine before using it.
Sodium citrate can not be used; an ultrafiltrate of plasma from
blood, which has been mixed with citrate in the usual way
causes a response of the gut corresponding to at least 50 p. c.
higher histamine concentration as compared wdth a heparinized
sample. Similar results are obtained if citrate mixed with a
known histamine solution is directly tested. The citrate itself
causes a contraction after which the histamine sensitivity of
the gut is diminished. Contrary to Minard (1941) we have found
that citrate affects the histamine values even if the sample has
gone through the chemical extraction procedure.

If the facts mentioned above are taken into consideration,
the ultrafiltration method seems to be well suited for the pre-
paration of plasma prior to the biological assay. To the chemical
extraction method it has the advantage of being less drastic;
if histamine is shown to be present in an ultrafiltrate of plasma,
it can not reasonably be said that the histamine has been pro-
duced e. g. from histidine which is claimed to occur during the
processes of chemical extraction (Akerblom 1941). The ultra-
filtration requires less work and supervision than the extraction
processes and it also takes less time; the extraction lasts about
three hours, the ultrafiltratioh only about one. Using a centrifuge
of higher speed and improving the method in other respects we
might be able to reduce the time required even more and thus
perhaps get the sample ready for assay during the course of the
actual experiments.

Chapter 3

Identification of the gut contracting substance
in ultrafiltrates of plasma

»Keine der biologischen Methoden zum Nachweis von Hista-
min in Gewehsextrakten ist so spezifisch wie einige der Proben
auf Acetylcholin» (Gaddmn 1936). It seems significant that
many authors avoid the term histamine and prefer to refer to
2 >histamine-like substances®, »histamine activity®, »H-substance»,
»histamine yielding substance® etc. A disadvantage of the ultra-
filtration method as compared with the extraction method is
that the ultrafiltrate may contain gut contracting substances
other than histamine. It thus seems imperative to ascertain to
what extent the gut contracting effect of the ultrafiltrate is
due to histamine. In the body there exist a number of substances
which can elicit contraction of the guinea-pig gut, e. g. acetyl-
choline, choline, vasopressin, kallikrein, P-substance, adenosine
triphosphate (Buchthal and Kahlson 1945), creatinine, tyramine,
potassium, citric acid. As far as the ultrafiltrate is concerned
some of these substances can be disregarded at once. Kallikrein
is not ultrafiltrable and besides it is believed to occur as an
inactive compound in the blood. Some of the substances are
not contained in plasma, in any case not in a concentration
high enough to cause contraction of the gut, e. g. acetylcholine
and potassium; in our control experiments choline, creatinine
and citric acid in plasma concentrations have not affected the
ileum; P-substance is said not to be present in blood (Zipf and
Hiilsmeyer 1933). Whether there is any tyramine in blood is
not definitely ascertained; Wolf and Heinsen (1935) have not
been able to find it there whereas Freund (1936) believes that

Fig. 1. Isolated guinea pig’s ileum in 2 ml
Tyrode solution. 1: O .020 y histamine. 2;

0.030 Y histamine. 3: O.s ml ultrafiltrate of
guinea-pig’s plasma. 4; O .020 y histamine.

5 and G: O .2 ml ultrafiltrate. 7 and 9: O.i ml
ultrafiltrate. 8: O.oio y histamine.

the »Spa,tgift» of the blood is identical with tyramine. In the
ultrafiltrate there may of course occur active substances which
are not yet identified; Barsoum and Gaddum (1935) e. g. de-
scribed an unknown gut contracting agent in blood.

In this chapter we have examined on some test objects well
suited for the identification of histamine if the gut contracting
agent of plasma ultrafiltrates qualitatively and quantitatively
corresponds to histamine. The samples have also been tried on
some test objects which are affected by some of the other gut
stimulating substances mentioned above but not by histamine
in concentrations which are likely to occur in plasma. Most of
the experiments described in this chapter have been made with
guinea-pig’s plasma, but plasma from cats, dogs and rabbits
has also been used.

1. Experiments on isolated guinea-pig's ileum. The active
substance of ultrafiltrate corresponds to histamine in the fol-
lowing respects:

a) the contraction elicited by ultrafiltrate is exactly similar to
that caused by histamine (see e. g. fig. 1 and 2). The contrac-
2

Fig. 2. Isolated guinea-pig’s gut. 1: O .020 y histamine. 2: O.i ml ultra-
filtrate of guinea-pig’s plasma. 3-: O .022 y histamine. 4: O.i ml extract
of the same plasma. 5; The hath contains 0.5 y thymoxiethyldiethyl-
amine per ml. 6; Washing of the gut with Tyrode solution for several
times. 7: The bath contains 0.5 mg theophyllinemonoethanolamine
(theamine) per ml. 8: O.i ml untreated plasma of the same guinea-pig.

tion takes place almost instantly after the injection, soon
reaching its maximum, then followed by relaxation after a
single washing.

b) the contractions caused by the ultrafiltrate and by histamine
follow the same concentration-action curve (fig. 1).

c) the two histamine antagonists thymoxyethyldiethylamine
(Bovet and Staub 1937, Minard and Rosenthal 1939) and
theophyllinemonoethanolamine (Emmelin, Kahlson and Lind-
strom 1941) reduce the sensitivity of the gut to ultrafiltrate and
histamine to the same extent (fig. 2).

d) Marthe Vogt (1943) investigating the point of attack of
some drugs on rabbit’s intestine found that if the gut was kept
in a refrigerator for some time the nervous apparatus was
destroyed prior to the muscle cells; at this state only drugs
directly affecting the muscle were active. During the experi-
ments on preserved guinea-pig’s gut we tried to find out if the

1 2 3 4 5

Fig. 3. Action of plasma ultrafiltrate A) on guinea-pig’s gut, kept in
refrigerator for 5 days and B) on the fresh gut. A. 1: O.oio y histamine.
2: O.030 j> histamine. 3: 0.2 ml ultrafiltrate of guinea-pig’s plasma.
4: 0.035 y histamine. 5: 0.3 y acetylcholine. B. 1: O.oi y acetylcholine.
2: 0.03.') y histamine. 3: 0.2 ml ultrafiltrate. 4: O.C30 y histamine.

specifity to histamine can be augmented in this way as it is
well-known that histamine stimulates the muscle cell directly
(Gasser 1926). It has, however, been difficult to define the
degree of the cold treatment when the histamine sensitivity is
still high but the sensitivity to e. g. acetylcholine is lost. Yet
figure 3 shows an experiment on a preserved gut, the acetyl-
choline sensitivity of which was greatly reduced; it is evident
that the effect of ultrafiltrate on this gut is of the histamine and
not of the acetylcholine type.

2. Experiments on blood pressure.

a) the blood pressure of the rabbit, anaesthetized with
urethane, is not depressed by ultrafiltrate. It is well known
that histamine in the concentrations in question does not affect
the rabbit’s blood pressure either.

b) like histamine ultrafiltrate depresses the blood pressure of
the anaesthetized and atropinized cat. The effect of ultra-

Fig. 4. Action of ultrafiltrate of guinea-pig’s plasma on A) cat’s blood
pressure, B) guinea-pig’s gut.

A. Blood pressure of cat (3.8 kg) under chloralose. The pressure in the
carotid artery was registered by means of a mercury manometer. Time

in minutes.

1: 0.30 y histamine, injected into a femoral vein. 2: l.o ml ultrafiltrate.
3: 0.25 r histamine. 4:. 0.23 y acetylcholine. Between the two sections of
the tracing there is an interval of about 2 hours. During this time
histamine was slowly injected intravenously in gradually increasing
doses, beginning with 0.5 y per kg of body weight per minute. The
injection of the highest dose, 50 y per kg per minute was interrupted 2
minutes before the second section of the tracing.

From the first part of the tracing it can be seen that the action of 1 ml
of ultrafiltrate corresponds to about 0.27 y histamine.

B. Guinea-pig’s ileum. 1: 0.Q25 y histamine. 2 and 4: O.l ml ultrafiltrate.
3: 0.020 y histamine. 5: 0 . 02 o / histamine. In this assay 1 ml ultrafiltrate

corresponds to about O .23 y histamine.

filtrate on cat’s blood pressure is shown in fig. 4. It is also
obvious from figure 4 that a fairly good quantitative agreement
is obtained with the two test objects, cat’s blood pressure and
guinea-pig’s ileum.

c) we have tried to work out a test specific for histamine on
cat’s blood pressure in the following way. In chapter 6 it will
be shown that by intravenous injection of histamine in gradually
increasing doses it is possible greatly to raise the histamine
content of cat’s plasma maintaining a normal blood pressure.
This procedure diminishes the sensitivity to histamine while

Fig. 5. Blood pressure in the femoral arterj' of cat (2.8 kg) under chlo-
ralose and artificial respiration. Time in minutes. 1: 0.25 y histamine.
2: O.so y acetylcholine. 3: 15 mg sodium glycocholate. At 4 a slow in-
jection of histamine was started. At the beginning 2.7 y histamine per
kg and minute was injected, then 10 y and 30 y per kg and minute, each
dose during 20 minutes. The injection was interrupted 3 minutes before
e section B. Between B and C there is an interval of 5 minutes.

5; 2.5 y histamine.

Other depressant agents are still strongly active. We have not
yet finally established the specifity of this test, but we have
ound that the depressant effect of two arbitrarily chosen sub-
stances, acetylcholine and glycocholic acid, is not reduced after
such histamine treatment. Fig. 5 shows that slow histamine
injection reduces the effect of histamine but not that of acetyl-
choline or the bile acid; in chapter 6 it will be seen that the
sensitivity to histamine is successivly restored when the hista-
mine content of the plasma decreases to normal.

As a test of specifity Barsoum and Gaddum (1935) used the fowl’s
ree a coecum bathed in a strong histamine solution; the gut was eon-
rac e m the usual way by certain substances but not by histamine,
e method described above seems to be of the same tj'pe as- that of
arsoum and Gaddum but we hope that with our method it will also
e possible to identify intravitally liberated histamine, e. g. in reactive
yperaemia. In preliminary experiments we have found that a prepara-
on 0 a certain snake venom, which exerts its depressing effect chiefly
y 1 erating histamine from the tissues is less active on the blood
pressure after slow injeetion of histamine.

We have made the observation that by the histamine pre-
reatment the effects of ultrafiltrate and histamine are reduced
to the same extent (fig. 4).

Fig. 6. Action of ultrafiltrate of guinea-pig’s plasma A) on cat’s blood
pressure, B) on guinea-pig’s gut;

A. Blood pressure in the carotid artery of a cat (2.9 kg) under chloralose
and artificial respiration. Reduced circulation. Injections were made
in the coeliac arterj' through a cannula, pointing towards the aorta.
1: 0.15 r histamine dissolved in 0.7 ml saline solution. 2: 0.7 ml ultra-
filtrate. 3 and 4: 0.7 ml saline solution. 5: 5 j" adrenaline • in 0.7 ml

saline solution.

In this assay the histamine content of the ultrafiltrate, somewhat exceeds
200 y of histamine per litre.

B. 1: 0.1 ml ultrafiltrate. 2: 0.oi5 y histamine. 3: 0.oi7 y histamine. 4: O.i

ml ultrafiltrate, 5: O .022 y histamine. 6: O.i ml ultrafiltrate.

Result of the assay on the gut: the ultrafiltrate contains about 200 y of

histamine per litre.

d) ultrafiltrate, like histamine, under special circumstances
raises the cat’s blood pressure. Dale (1920) showed that
histamine can stimulate the adrenaline secretion from the
adrenal medulla, Feldberg (see e. g. Feldberg and O’Connor
1937) made use of this fact for the identification of histamine.
Quantities as small as O. 05 — O .3 y of histamine may be detected
in this way if the circulation of the cat is reduced by eviscera-
tion and histamine injected in the arterial blood stream close to
the adrenals (Szczygielski 1932).

We performed such experiments as follow's. In cats under chloralose
the %'agi were cut. V. portae, aa. mesenter. sup. et inf., a. coeliaca
about one cm from aorta, were ligated. Cardia and rectum, the kidney
vessels at hilus, aorta and v. cava inf. below the adrenals were also
tied. A cannula was placed in the central stump of the coeliac artery,
and histamine and ultrafiltrate injected.

Fig. 7. Ultrafiltrate of guinea-pig’s plasma, tested on perfused guinea-
pig’s bronchi (A) and isolated guinea-pig’s gut (B). In a guinea-pig
under ether bloodi was withdrawn by heart-puncture. A plasma sample
was ultrafiltrated. The animal was then anaesthetized with urethane
subcutaneously; the bronchi were perfused and a piece of gut isolated.

A. Isolated bronchi. Bronchoconstriction is indicated by larger excur-
sions of the volume recorder. Time in minutes. 1: O .025 y histamine.

2: 0.3 ml ultrafiltrate. B: O.ooo y histamine.

According to this estimation the liistamine concentration of the ultra-
filtrate is something between 100 and 150 y per litre.

B. Isolated gut. 1: O.oio y histamine. 2 and 4: O.ois y histamine. 3: O.i ml
ultrafiltrate. 5: O .020 y histamine. C: O .017 y histamine. Result: the ultra-
filtrate contains about 160 y of histamine per litre.

Fig. 6 shows that ultrafiltrate like histamine raises the blood
pressure under these conditions. The results also show a fairly
good agreement with the assay on the gut.

3. Experiments on the isolated guinea-pig’s bronchi. Feldberg
et al. (1932) found that histamine causes a bronchoconstriction
in the guinea-pig’s lung perfused with saline. We have tried
to make use of this preparation for estimation of histamine.

The bronchial tone of a guinea-pig under urethane was registered
according to Konzett and Rossler (1940). This method of registration
as used in this laboratory has already been described (Emmelin, Kahlson
and Wicksell 1941). In the following experiments the chest was opened
and a cannula inserted through the right ventricle into the pulmonary
arterj'. Oxygenated Tj'rode solution of body temperature ran from a
pressure bottle through the cannula. Injections of histamine and ultra-
filtrate were made through the rubber tube close to the cannula. The
perfusate ran away through another cannula in the left auricle. We
have found that an injection of lieparine to the animal before the pre-

paration greatly improves the method as clotting in the lung vessels is
thus avoided. Unfortunately the registration of the bronchial tone is
affected by an oedema of the lung developing during the perfusion,
especially after injection of histamine. In an attempt to counteract this
oedema we have added »dextran> (Grdnwall et al. 1945) in a suitable
concentration to the Tyrode solution.

The advantage of this test object is its high histamine sensitivity.

From figure 7 it is apparent that the ultrafiltrate like hista-
mine constricts the guinea-pig’s bronchi. The figures obtained
are in rather good accordance with those obtained on guinea-
pig’s ileum.

4. Experiments on other test objects. Experiments have also
been made on isolated rabbit’s heart in Langendorff’s arrange-
ment, on frog’s heart beating on a Straub’s cannula, on the
isolated rectus muscle and on strips of stomach of the frog.
Neither ultrafiltrate nor histamine in moderate concentrations
have any significant effects on these test objects.

The experiments so far referred to have shown that the
biological effects of a plasma ultrafiltrate are identical with
those of histamine. On testing the same ultrafiltrate on dif-
ferent objects a good quantitative agreement has been obtained,
a fact which suggests that the biological effects are due to one
and the same substance and that this substance is identical
with histamine.

Chapter 4

Is all the blood histamine contained in the
corpuscles?

It is well known that shed blood easily aquires a biological
activity which is not to be found in the circulating blood. This
can be induced by trauma, clotting, or simply by storing of the
blood. In fresh serum there is an agent which affects blood
pressure and smooth muscle; Freund (1920) called this agent
»Fruhgift» and Zipf (1931) identified it as adenylic acid. It is
rapidly destroyed enzymatically, and the serum then causes
other effects, due to »Spatgifte» ()S-thrombovasin according to
Simon 1938, Simon and Komlos 1939); Freund (1936) suggests
that these »Spatgift»-effects are caused by tyramine which,
however, is denied by Simon (1937). These unidentified sub-
stances are believed by the authors quoted above, normally to
be present in the platelets; they are liberated when the platelets
are destroyed by shaking, by clotting or by mere storing of
the blood. Gut contracting agents have recently been found in
serum and platelets by Zucker (1944) and Tsai, Me Bride and
Zucker (1944). Many investigators have observed that blood,
injured by shaking, hemolysis or standing can cause circulatory
effects (Phemister and Handy 1927, Feldberg, Flatow and Schilf
1929, Newton 1932, Fleisch 1937). Shaking or clotting augments
the adenosine activity of blood (Barsoum and Gaddum 1935).
The substances discussed here may, of course, interfere with
our tests.

Histamine is of special interest in this connexion. O’Connor
(1912) found that serum has certain effects on smooth muscles
which plasma has not. He suggested that the active agent

originated from the platelets and also stressed that its effects
were very similar to those of the recently discovered ^-imida-
zolylethylamine. Schwartz (1936) found that after clotting of
the rabbit’s blood histamine appears in the serum and he
believed that the histamine originated from the platelets; he
also suggested that shaking liberates histamine. Barsoum and
Smirk (1936) showed that the histamine content of plasma
increased if the blood was kept for some time before removing
the corpuscles by centrifugation. According to Code (1937)
most of the blood histamine can be recovered from serum. Code
also observed that if clotting is prevented by potassium oxalate
and the blood is left standing for some time the histamine
concentration in plasma is increased.

Taking all these observations into account, and also the fact
that most of the blood histamine, at least in the rabbit, is
contained in the extremely fragile platelets it is tempting to
assume that all the histamine, which is found in a plasma
sample, originates from the platelets of the circulating blood.
It must be born in mind that trauma and clotting, which are
known to liberate active substances, are at work during the
procedures involved in the securing of a sample of blood
plasma. O’Connor (1912) pointed out the difficulties of avoiding
injuries to blood cells at the blood collection. Minard (1941),
who found that of the total blood histamine in the rabbit only
2 — 3 per cent was present in the plasma, suggested that the
plasma histamine might have been liberated from platelets
destroyed outside the body. Barsoum and Smirk (1936) also
discussed the possibility that histamine detected in a plasma
sample is liberated from the blood cells in vitro.

In a series of experiments we have tried to find out to what
extent a rough handling of the blood affects the plasma hista-
mine values. We have also investigated the effect of clotting,
since Anrep et al. (1939), contrary to Code, claim, that on
clotting, histamine does not appear in the serum but can be
extracted from the fibrin fraction. In a second series of expe-
riments we have tried to establish whether or not histamine is
a normal constituent of plasma.

1. The effect of hemolysis. In a preliminary experiment we

obtained a plasma sample rich in platelets in the following way.
From an ear vein of a rabbit blood was collected in a ghiss
tube containing heparine and was then slowly centrifuged for
a short time (1000 rev./min. during 15 min.). A sample of the
turbid plasma was withdrawn and the blood then centrifuged
at a higher speed until the plasma became clear (3500 rev./min.
during 30 min.). A five-fold volume of distilled water was added
to the turbid plasma and the sample was ultrafiltrated. The
clear plasma was also ultrafiltrated. The assay on the guinea-
pig’s gut showed that the turbid plasma contained about 2000 y
of histamine per litre, the clear plasma 185 y/l. It is obvious
that the lysis of the corpuscles is sufficient to liberate histamine;
since the ultrafiltrate contained physiologically active histamine,
this histamine must have been present in the corpuscles in an
active state or as a very labile inactive compound,

2. The effect of shahing. A sample of heparinized rabbit’s
blood was divided in two portions. One was centrifuged im-
mediately, the other after ha%dng been moderately shaken a
few times in the tube. The e.xperiments were carried out with
blood from three rabbits. The results of these experiments are
given in table 1, where plasma A is obtained from the unshaken,
plasma B from the shaken sample. Tlie plasma samples were
extracted (experiment nr 1 and 2) or ultrafiltrated (experiment
nr 3) and the histamine content was determined. It is obvious
that in rabbits the plasma histamine concentration increases on
shaking.

TABLE 1,

Experi-

ment

Hi content, f/l

plasma A

plasma B

blood

840

1500

2500

200

2200

1950

150

600

—

3. The effect of clotting. From the ear vein of rabbits, blood
was collected alternately in two centrifuging tubes, one of which

contained heparine. The samples were left standing for a few
hours at room-temperature. The heparinized blood remained
liquid whereas the other soon clotted and serum and clot began
to separate. The samples w’ere centrifuged; the clear fluids were
collected and the histamine content estimated after ultrafiltra-
tion. Table 2 indicates that in our experiments the histamine
content of rabbit’s serum is considerably larger than that of
plasma.

TABLE 2.

Experiment nr

Hi content

plasma

800

630

250

560

500

150

Y/1

serum

4000

3000

1100

3000

2500

There is every reason to assume that at least most of the
gut activity of serum is due to histamine: the effect of serum
ultrafiltrate is antagonized by thymoxyethyldiethylamine; the
ultrafiltrate also depresses the blood pressure of the atropinized
cat to an extent which in terms of histamine corresponds to
the activity on the gut. Besides, histamine has been chemically
identified in the »white layer» of blood from the rabbit (Code
and Ing 1937).

4. The histamine content of plasma from the same individual
over a period. It is well known that the histamine concentration
of blood, while warying within fairly wide limits in a species,
remains rather constant during a long period in one and the
same individual (Code and Mac Donald 1937, Zon, Ceder and
Crigler 1939, Rose and Browne 1940, Minard 1941). If the
histamine detected in a plasma sample in vitro originates from
the corpuscles it seems likely that subsequent plasma samples
from one individual should — contrary to the blood samples —
show great variations in histamine content owing to greater
or smaller injury to the blood on different occasions. During
a period we have estimated the histamine concentration of
plasma and of blood in a number of animals. Hypothesing that
histamine occurs in the circulating plasma nothing is known

as to whether it occurs there in constant amounts as is the
case in total blood. We have investig-ated to what extent the
histamine concentration of other tissues than blood is maintained
constant; excised pieces of skin have been examined. Tlie
experiments were carried out on guinea-pigs, rats, dogs and
rabbits.

a. Experiments on gitinea-pigs. The figures from 9 animals
are summarized in table 3. The body weight of each animal
at the beginning of the experiment is also given. The blood
was collected from the etherized animal by heart puncture; 5 — 9
ml of blood was withdrawn in a syringe, containing heparine.
1 — 2 ml of blood wms extracted. The remaining blood w’as
centrifuged at 3500 rev. per min. for 15 — 30 minutes and plasma
was collected and extracted. An area of the abdominal skin
was carefully shaven and a piece of skin excised, after wdiich
the wound was closed with sutures. The subcutaneous tissue
Avas carefully removed from the excised skin which Avas then
weighed hs soon as possible to prevent loss of weight due to
evaporation of Avater (the skin samples usually weighed about
100 mg). The skin was then minced Avith scissors and ground
in a mortar with quartz powder and trichloracetic acid for 30
minutes. After centrifugation the sample Avas extracted accord-
ing to Code. Samples of plasma, blood and skin Avere assaj^ed
on guinea-pig’s ileum.

It is obvious from the table that the histamine concentration
not only of blood and skin but also of plasma is kept at a fairly
constant level in an animal for Aveeks and months. Considering
that in these experiments the plasma histamine is sustained at
a rather constant level in one and the same individual, it is
not unlikely that the circulating plasma really contains hista-
mine; and further, it is tempting to assume that some regulating
mechanism endeavours to keep the histamine concentration of
plasma, like that of total blood and skin, at a constant level.
This explanation seems to us more reasonable than e. g. the
assumption that the blood collection should ahvays cause the
same degree of damage and that the cells should possess a
fragility, typical to each individual but varying Avithin Avide
ranges in- a species.

TABLE 3.

Guinea-pig nr

Date of

Histamine content of

ana

body-weight

experiment

plasma, f/l

Blood, f/1

Skin, mg/kg

2G.2

200

615

7.3

220

600

10.4

180

580

26.4

225

600

8.5

200

570

1.3

200

700

6.5

14.3

225

—

5.7

635

27.3

—

750

7.4

6.4

210

700

—

1.3

135

430

2.5

7pn

13.3

150

400

1.9

9.4

170

410

2.3

3.3

325

27.3

335

4.5

840

10.4

290

5.8

24.4

—

6.2

5.3 ,

125

450

27.3 '

125

500

4.8

11.4

—

400

3.8

30.5

135

520

—

26.7

130

*—

—

6.3

250

475

11.4

AQf)

26.3

265

450

13.9

Dl7v/

17.4

—

450

—

14.3

375

3.3

28.3

150

340

2.8

690

12.4

130

325

3.1

27.4

150

360

—

15.3

320

670

12.5

460

10.4

300

650

14.9

16.3

225

28.3

—

11.3

16.4

250

14.7

b. Experiments on rats. 7 big white rats were used for these
experiments which were carried out similar to those on guinea-
pigs. 3 — 6 ml of blood was obtained by heart puncture; the

blood was used for estimation of histamine in plasma. The skin
samples weighed 25—75 mg. The results, summari/.ed in table
4, agree with those of table 3.

TABU-: 4.

Rnl nr

Date

].9.4

3.r.

30..')

2.').-}

1 7 .r>

10.4

21)

24.5

Plapma Hi, y/1

140

100

2,50

22o

ir.o

165

150

Skin Hi, yfl

24.8

27.9

22.3

,30.0

32.0

35..5

io.r»

1G.5

20.5

Rat m

Date

18.4

17.:.

20.4

10.4

17.,5

17.4

IQI

l.G

Plasma Hi, '(l\

110

100

190

300

145

1G5

Skin Hi, yP

10.0

ir..o

31.9

27.0

c. Experiments on cloys, Tlie Justaminc content of plasma
was determined in two (log.s. In an ear vein of a nonannesthc-
tized dog an incision was made from which about 20 ml of
blood was collected in a centrifuging tube, containing heparine.
The sample was treated as in the previous experiments. Table 5
demonstrates the constancy of the plasma values.

TABLE 0.

Dog nr

Date

20.2

IQI

15.3

7.4

27 2

6.3

8.3

18.5

Plasma Hi, y/l

d. Experiments on rabbits. 10 — 15 ml of blood was collected
from an ear vein of the nonanaesthetized animal. From table G
It is obvious that the histamine concentrations of the plasma
samples show- very great variations while the histamine content
of blood remains fairly constant.

In this respect the rabbit obviously differs from the other

TABLE 6.

Babbit nr

' 1

Date

9.2

19.2

r3.3

18.4

25.5

30.5

10.3

26.3

12.4

23.4

Plasma Hi, f/l

570

800

600

200

380

390

1900

140

280

Blood Hi, f/l

2500

3000

—

2800

3800

—

4000

Babbit nr

Date

18.1

23.1

24.1

29.4

3.5

3.6

9.6

20.3

29 5

11.7

Plasma Hi, yA

170

800

150

1200

300

200

150

750

210

250

Blood Hi, y/1

4200

4500

3300

2800

—

3800

4200

species investigated. It may be pointed out that in a paper by
Chute and Waters (1941), not directly concerned with the
subject under discussion, we have found very great variations
of the plasma histamine in one and the same rabbit. In rabbits
the platelets have an exceptionally high histamine content and
liberate histamine very easily in vitro. It is reasonable to
assume that the histamine, which can be extracted from the
rabbit’s plasma, mainly originates from the platelets. We have
investigated the plasma histamine in experiments where the
blood was handled as carefully as possible. It may be mentioned
that the plasma values in table 6 are of the same order as
those given by other authors (e. g. Zon, Ceder and Crigler 1939,
Chute and Waters 1941, Rocha a Silva and Andrade 1943).
Table 7 summarizes 6 experiments of this kind. The first
plasma sample (1) was obtained in the usual way, the second
(sample 2) as follows. The blood was collected from ear veins
of nonanaestetized animals (rabbits nr 7, 8 and 9) or from the
carotid artery of animals under urethane (rabbits nr 10, 11
and 12). The blood was collected in a paraffined centrifuging
tube containing heparine and cooled in ice water. The sample
was centrifuged. In some cases the tube was placed in a larger
centrifuging tube, containing ice water in order to cool the
sample during the centrifugation. The tube was paraffined as
this is known to protect the platelets. The blood was cooled

because heating may release histamine from cells and cellular
debris (Trethewie 1938).

TABLE 7.

Rabbit nr

Hi content

sample 1

330

340

450

840

350

750

t/1

sample 11

135

130

210

The table shows that plasma obtained under these conditions
has a relatively low histaihine content. When using a similar
method, Minard (1941) also found a low histamine concentra-
tion in rabbit’s plasma. In experiments with blood from the
rabbit it is obviously of great importance to protect the cor-
puscles from injuries. In experiments with other species the
plasma histamine does not decrease when special precautions
are taken to protect the corpuscles. Only in rabbits we have
investigated blood from one and the same animal, treated in
the two different ways; it may, however, be mentioned that the
average plasma histamine content of 14 guinea-pigs, the blood
of which had been obtained in the usual way, was 191 y per
litre, while the average of 23 animals, the blood of wdiich had
been treated according to the more careful method, was 196 y
per litre.

These experiments show that the plasma histamine in guinea-
pigs, rats and dogs, as determined by these methods, is kept
fairly constant in the same individual and they indicate that
histamine is a normal constituent of plasma. This is supported
by the fact that histamine is present in aqueous humour (Emme-
lin and Palm 1944; further experiments are described in chapter
5). Even if the conception that the aqueous humour is produced
from plasma by an ultrafiltration process is not accepted by
all investigators, there is no reason to suppose that the hista-
mine present in the aqueous is not derived from the plasma.
A similar assumption was made by Barsoum and Smirk (1936),
who detected histamine in oedema fluid from patients with
congestive heart failure; from this fact they concluded that the
3

histamine found by them in plasma of these patients had not
leaked out of the corpuscles in vitro.

Contrary to the other species investigated the rabbit shows
great variations in the plasma histamine concentration of one
and the same individual. In this animal the plasma histamine
can be greatly decreased by careful handling of the blood.
The question arises whether rabbit’s plasma normally contains
any histamine at all. Some facts indicate that this is the case.
In rabbits the aqueous humour contains histamine. By careful
handling of the blood, histamine values are obtained which
are almost as low as those of aqueous humour. In chapter 5
it will be shown that by using a special technique — ultra-
filtration in vivo — histamine values of the rabbit’s plasma
are found which are just as low as those of aqueous humour —
but not lower.

Chapter 5

Is the plasma histamine present in a physio-
logically active or inactive form?

Several authors have stressed that using the common methods
of extraction it is not possible to decide whether histamine
detected with the biological test objects was really present in
the circulating blood in a free, active form (Gaddum 1936,
Barsoum and Smirk 1936, Feldberg 1937, Dale 1937 — 38, Code
and Mac Donald 1937). It should be remembered that the
biological assay is proceeded by drastic extraction processes
during which the sample i. a. is boiled for 1 hours in strong
hydrochloric acid. It has even been suggested that the hista-
mine found is derived from histidine, which has been decarboxy-
lated during the extraction (Akerblom 1941). Dale (1936)
suggests that the plasma histamine may, like kallikrein, form
an inactive compound with some plasma component and Holtz
(1937) presents a similar hypothesis. Kaiser (1939) is of the
opinion that the plasma histamine is physiologically inactive
and probably adsorbed to protein.

It is a priori not improbable that histamine is present in
plasma as a physiologically inactive compound. It is known
that histamine can occur in a conjugated form in the body. In
the urine Anrep et al. (1944) found a compound from which
histamine could be released by acid hydrolysis. It is not ji-et
clear in which state histamine is present in the cells but many
investigators suggest that it exists as some inactive compound.
Rocha e Silva (1943) is of the opinion that the histamine of
the cells is bound to protein in a peptide linkage, the proximal
amino acid in the chain being either arginine or lysine.

As plasma can not, without previous treatment, be tested on
the usual test objects we have approached the problem under
discussion by using a preparation method which is not so drastic
as the chemical method; we have applied the ultrafiltration
method. Even in this process histamine might be liberated from
a hypothetic, very labile compound, and we have therefore
also tackled the problem in other ways.

1. Ultrafiltration of plasma samples.

In these experiments a plasma sample was divided into two
portions, one of which was ultrafiltrated, the other extracted
according to Code. The extracts were dissolved in distilled
water to the same volume as that of the original plasma sample.
The histamine values are given in terms of y per litre ultra-
filtrate or extract. The assay was made on guinea-pig’s gut.
Some samples have for the purpose of identification been
treated according to the methods of chapter 3. Some of the
results given here have already been published in a preliminary
report (1945).

a. Experiments on guinea-pigs. The blood was obtained from
the etherized animal by heart puncture as described in chapter
4. As the plasma sample had to be large enough for both
extraction and ultrafiltration a somewhat greater amount of
blood (10 — 15 ml) was usually withdrawn. The results of 38
experiments are shown in table 8.

b. Experiments on rats. Blood was collected as in chapter 4,
The results of 12 experiments are given in table 9.

c. Experiments on cats. On cats under chloralose or ether

15 — 30 ml of blood was withdrawn through a cannula in the
femoral artery. The blood was collected in a cooled, paraffined
centrifuging tube containing heparine. Since it has been stated
that the histamine content of blood is increased at the be-
ginning of a chloralose narcosis (Kwiatkowski 1943) the blood
samples were taken 1 — 2 hours after the injection of chlora-

lose. Table 10 shows the results of experiments on 14 cats.

d. Experiments on dogs. The blood was obtained from the
ear vein of unanaesthetized dogs (nr 1 and 2 of table 11) or

Gtiitica-pijr nr

content, Y'J extr.

Guinea-pi!' nr

■1

250 2f)0

|l40

1.50|

1.50

2-JO - 100 1 -JO

l.i|ir>|lt; 17 18 If) 20 21

Histamine U. f. 270 2r,0|l30|l40|l30jJ5f)

content, Y/i cxlr. 2G0 — jllwInO

iBlD

32.'>

33.')

Ilf)

2.3.G!21o!j7o)nr)!llO

28 2!) 30 31 .32 33

Gninea-pig nr

10 11 12 1314

content, y/I extr. fiO

asm

Histamine V. f. 50 30 35 80 35 25 -JO 75 -15 OoIiuoIgO -15 50

from the femoral artery of animals under morphine-chloralose.
The results are summarized in table 11-
e. Experiments on rabbits. The blood was obtained from the
ear vein of unanaesthetized rabbits or from the carotid artery
of rabbits under urethane or chloralose. Table 12 shows 28
experiments on rabbits.

TABLE 12.

Rabbit

Histamine

U. f.

620

145

2000

1900

810

800

560

1400

310

175

content, y/l

extr.

2000

800

570

1200

350

150

Rabbit nr

Histamine

U. f.

800

170

150

200

280

185

800

300

420

330

content, f /I

extr.

800

—

150

200

—

200

840

310

390

330

Rabbit

—

Histamine

U.f.

360

130

200

250

150

—

content, y/1

extr.

380

130

210

250

135

—

In experiments nr 21 — 28 of table 12 the same precautions
were taken as in chapter 4 at the blood collection.

It is obvious from the tables that the ultrafiltration and the
extraction methods give approximately similar figures. (In the
case of cat there are some exceptions, the reason for which is
obscure; the ultrafiltrate in these instances ma}'' contain some
agent interfering with the assay.) We have not gone into a
detailed study of the concordance of the two methods of
preparing plasma samples and the discrepancies observed,
since this does not touch our primary problem. Our experi-
ments are designed to show that plasma contains histamine in
an ultrafiltrable form; the amount of ultrafiltrable histamine
corresponds fairly well to that which can be obtained by the
chemical extraction method. This fact obviously excludes the
possibility that the histamine is produced e. g. by decarboxyla-

tion of histidine at the chemical treatment of the plasma. On
the other hand these experiments do not definitely prove that
histamine is present in the circulating plasma in a free, active
state. There remains the possibility that the histamine occurs
physiologically in an adsorbed form or as an extremely labile
compound and is liberated during blood collection, centrifuga-
tion of the blood or ultrafiltration. It is, however, unlikely
that histamine is gradually liberated during ultrafiltration since
the histamine concentration of the filtrate is the same at the
beginning as at the end of the filtration. In a series of experi-
ments to be described later we have obtained plasma samples
for biological assay avoiding by special measures the trauma
of blood collection and centrifugation.

2. The histamine content of aqueoits humour.

The mode of production of aqueous humour still remains
controversial in essential points (for references see e. g.
Hodgson 1938, Davson 1939, Robertson and Williams 1939,
Davson and Quilliam 1940, Duke-Elder and Davson 1943,
Friedenwald 1944). Many investigators consider the aqueous
to be an ultrafiltrate of plasma. It must be of special interest
to study the histamine concentration of this filtrate, produced
within the organism, where the blood is not subjected to the
treatment which is necessary for the preparation of an ultra-
filtrate in vitro. It has already been shown that the aqueous
can be tested directly on guinea-pig’s gut without any previous
treatment (Emmelin and Palm 1944). The aqueous humour is
usually slightly alkaline as compared to the Tyrode solution
used, but we have observed that neutralization does not alter
the histamine values found, and we have therefore omitted this
procedure. The aqueous was obtained by puncturing the
anterior chamber of the eye with a fine needle. In many cases
the aqueous was injected into the test bath immediately after
the aspiration, and its histamine concentration determined.
When a plasma sample of the same animal was tested for
wmparison the aqueous was kept in a refrigerator for a few
hours. Usually the aqueous from both eyes were mixed.

a. Experiments on guinea-pigs. Aqueous humour and blood
were obtained from etherized guinea-pigs. The total amount
of aqueous from one animal was only about O.i ml and it must
be emphasized that the histamine values obtained must be con-
sidered as very approximate since the available quantity was
only sufficient for one test on the gut. Table 13 shows the
results of 24 experiments on guinea-pigs.

TABLE 13.

Guinea-pig nr

Histamine

content, yd

aqueous

200

250

175

200

150

125

300

190

150

110

plasma U. f.

190

250

175

160

165

330

240

plasma extr.

175

230

165

155

320

250

^9j

Guinea-pig nr

•15

Histamine

content, y/l

aqueous

12.0

200

140

160

250

175

250

300

225

250

plasma U. f.

185

175

130

140

220

250

175

200

225

plasma extr.

170

160

175

235

b. Experiments on rats. Also on rats the samples were
collected from the etherized animals. From both eyes only a
total of 0.O3 — 0.05 ml of aqueous could be obtained. It is obvious
that the assay with such small quantities yields only very
approximate values. Besides, a specially histamine sensitive
piece of gut was required and we have therefore only been
able to assay a few of the samples taken. Table 14 summarizes
the results of 6 experiments.

TABLE 14.

Eat nr

Histamine

content, y/i

aqueous

350

200

130

275

100

200

plasma U. f.

175

110

330

150

loO

plasma extr.

310

155

310

145

150

isolated 4: 0-«®^

3 ml oi Tats aq iitetatatoe. extietivel

. „ tiiece 01 w ..

ioUiea nuro»“*:

3il oi Tats aq ^ifetatawe. ^ extietoely

8 t“"*

„e to Ustam«e («P lesulte

Eiferime»<» o® „v,tatoe4 fro® J . ...

1 nl aQttooos trWeu tit tal^lo • ^^g^s inves

=n 1 -®'^

fxjjeriwents on d 9 about O.s iu4 p^Derituents.

.j'Tom one eye of ® 4^ ® „, tM®® under

A Table 16 . Wp luvestigalea ‘ ^ oi

amed. on rabbits. fielded ^ v oi

:"-%= .S.S -r; =*•’»£

V®' *e trt'Ser r*ttt *® “'fstom IW

Mma 200 r P® 210 r f,” f tff tW Olood ®®»?^®

fi6“r®r '«“■■ ®tte are found to f®“®/ „ibed in oW^' *’

remaining 5 raP®® ® precautions desrar
«,e ooilected t®® P

. .a. already been pointed

the 'bistamine ^''“^^^^^y^guteea-pigs. 1^^!" and

especially » histamine histamine con-

indicate that ^ order. In rabbits o{

plasma is of the sam ^^^giaierably smaller

oentration oS latoed on *o break-down of

ptema. TMo 0“^; „i the Mood a certam brea^^^

oJ very oarefu tbe . the histamine

pUtelets occurs. K about

aqueous sboul g ^.^^^^^ting plasma. Vernon-

ccncentiatron rn motive Mstnin>

The fact that mo n _ j, fiuin P

strated in ^ , by a process jain histamine

plasma within aqueous and plasma the view

Nitration -and that the^^^ concentration support

in approntmately ptasma in a than

that histamine is p vm-wever, that other p .

should be remerAoied,^ the protoc«

ultrafiltratmn mrg vacations that se the

aqueous. There a ^ ".resins! a membrane

rjS«p"-eiultrafiltrateofpiasmau^^

of defined permeability-

3. UUrafiltration in vivo.

In these experiments we used the ordinary arrangement for
ultrafiltration in vitro. A stopper was placed in the opening
of the cellophane tube and through it two thin glass tubes
were inserted, a short and a long one reaching almost to the
bottom of the cellophane tube. The glass tubes were connected
with cannulae by means of short rubber tubings. The first
experiments were carried out on dogs under morphine-chlora-
lose. In these experiments heparine was injected intravenously
to prevent clotting. The femoral artery was exposed in two
places on the thigh. In the central part of the artery one of
the cannulae of the ultrafiltration system was inserted and
blood was allowed to fill the system. A clip was then put on
the artery centrally to the cannula and left there for the short
time during which the other cannula was inserted peripherally
in the artery. When the preparation was ready the clip was
removed and the blood flowed through the leg passing the
filtration system. Ultrafiltration was thus carried out, the
blood pressure of the animal acting as filtration pressure. A
clear, colour-less fluid accumulated in the centrifuging tube.
To prevent evaporation the opening of the tube was closed
mth paraffin.

TABLE 16.

Dog nr

Ultrafiltration in vivo

Histamine content, y/1

time of
collection
hours

volume

histamine

content

T/1

aqueous

plasma

U. f.

plasma

extr.

0.2

0.4

0.5

1.3

Experiments on 4 dogs are summarized in table 16. The
IS amine concentration was determined in in-vivo-ultrafiltrate,

aqueous humour and in ultrafiltrate and extract of plasma from
shed blood.

It seemed of special interest to experiment on rabbits with
this technique because of the discrepancy between the histamine
values of aqueous humour and ultrafiltrate of plasma. In these
experiments the ultrafiltration system was connected with the
carotid artery of animals under urethane or chloralose. Table
17 shows 5 experiments on rabbit. Fig. 9 illustrates some details
of the assay in experiment nr 2 of table 17.

TABLE 17.

Rabbit nr

Ultrafiltration in vivo

Histamine content, y/I

time of
collection
hours

volume

histamine

content

T/l

aqueous

plasma
U. f.

plasma

extr.

0.2

0.4

150

135

0.4

0.9

135

120

1.1

290

280

From the two tables it is evident that the volume of ultra-
filtrate obtained varied considerably. In the first experiment
on dog and rabbit a small cellophane tube was used, in the
following experiments a larger one. As the blood pressure was
registered only in a few of these experiments, we can give no
information of the part played by this factor.

In these experiments an ultrafiltrate of blood plasma was
secured with a minimum of damage to the blood. It is of interest
to note that in rabbits the ultrafiltration in vivo yields a fluid,
the histamine concentration of which is smaller than that of
the ordinary plasma ultrafiltrate and equal to that of the
aqueous humour.

The presence of biologically active histamine in in-vivo-
ultrafiltrates in a concentration equal to that of aqueous and

Fig. 9. Isolated guinea-pig’s gut. 1: O.oio y histamine. 2 and 4: 0.2 ml of
in-vivo-ultrafiltrate from rabbit’s blood. 3: O.oi4 y histamine. 5: O .012 y
histamine. 6; O .2 ml of aqueous from the same rabbit. 7: O .012 y histamine.

in dog also to that of plasma gives additional support to the
conception that histamine is present in plasma in an active

state, or if not, as a compound which extremely easily liberates
active histamine.

4. Cross circulation ex'periments.

Still another mode of attacking our problem can be founded
on the fact, demonstrated in chapter 4, that the histamine
concentration of plasma is kept at a fairly constant level in one
an the same individual but varies greatly within the species.

us amine is present in plasma in an active form it should
e possible to elicit typical histamine effects in an animal with
ow plasma histamine concentration by transfusion of blood
high histamine concentration. On this
amp ion we have performed cross circulation experiments in

guinea-pigs. These animals are for several reasons well suited
for experiments of this kind. It is fairly easy to determine the
plasma histamine content in a great number of these animals.
Further, there is a tissue to be found in guinea-pigs highly
suitable as a test object for histamine, the bronchi, which is
both sensitive and well adapted to this special type of experi-
ment because it gives — contrary e, g. to the cat’s blood
pressure — a marked reaction to a rather slowly occurring
increase of the blood histamine concentration. This difference
between the bronchi and the blood pressure as test objects for
histamine has been pointed out by Emmelin, Kahlson and
Wicksell (1941).

The experiments were carried out in the following way.
Two selected guinea-pigs were anaesthetized with urethane
subcutaneousl5^ The bronchial tone of both animals was
registered according to Konzett and Rbssler (1940). Heparine
was injected intravenously. In the first experiments the cross
circulation was arranged as follows. Two cannulae were con-
nected by a short piece of rubber tubing obstructed by a clip.
This system was filled with Tyrode solution containing hepa-
rine. One cannula was inserted in the central stump of the
carotid artery of one animal, the other in the jugular vein of
the other, the cannula pointing towards the heart — and vice
versa. When the clips were simultaneously removed a cross
circulation was thus established, blood flowing from the carotid
artery to the jugular vein of the other animal. A disadvantage
of this method is that a large arteriovenous anastomosis is
formed and the capillary net thus deprived of an amount of
arterial blood. Considering the short duration of the experi-
ment this disadvantage might be of no consequence, but in
later experiments another technique was used for safety. On
each animal the longest possible section of the carotid artery
was freed and two cannulae, connected by a piece of rubber
tubing inserted, one pointing centrally, the other peripherally
thus allowing blood to flow through the system to the periphery.
Each of these cannulae had a side tube. By means of a short
piece of rubber tubing the central cannula of one animal was
connected with the peripheral cannula of the other. The whole

Fig. 10. Cross circulation experiment (second method). From above:
bronchial tone of guinea-pig A and of guinea-pig B, time in minutes and
signal marks. Between the two sections of the tracing there is an interval
of 8 minutes. 1: cross circulation continuing. 2: 0.5 y histamine injected
into a jugular vein of A. 3: the same dose injected in animal B.
Guinea-pig A: body-weight 360 g, histamine content of plasma 120 y/l.
Guinea-pig B: body-weight 340 g, histamine content of plasma 310 y/1.
Narcosis: 2 g urethane per kg of body-weight O.i g of sodium phenyl-
ethyl barbituric acid subcutaneously.

system was filled with Tyrode solution containing heparine.
Cross circulation was established at the moment when the clips
were transferred from the tubings connecting the animals to
those connecting the central and peripheral cannulae of each
animal. The blood then flowed from one animal through the
central cannula, the connecting tubing and the peripheral can-
nula of the other animal out into the branches of the carotid
artery. After some minutes the cross circulation was inter-
rupted. The histamine sensitivity of the bronchi was tested in
each animal by intravenous injection of histamine. Unfortuna-
tely, in these experiments the anaesthesia, in order to prevent

spontaneous breathing, must be very deep, which reduces the
histamine sensitivity.

In these experiments the plasma histamine concentration of
about one hundred guinea-pigs was determined. The blood was
withdrawn by heart puncture and plasma was ultrafiltrated
previous to the assay on the gut. Two animals, one with a very
low, the other with a very high plasma histamine value, but
of about the same bodj’- weight were selected for each experi-
ment. We found 9 such pairs having the following plasma
histamine concentrations: 100 — ^335, 120 — ^270, 125 — 275, 130 —
300, 100—380, 175—265, 140—230, 140—350, 120—310 y per
litre. The actual experiments were always performed more
than a week after the heart puncture.

In 6 out of the 9 experiments a marked bronchoconstriction
was observed in the animal with the lower plasma histamine
content. In 3 experiments no bronchoconstriction was elicited
(the pairs 2, 3 and 4). In these 3 experiments the histamine
sensitivity of the bronchi of the animal with the smaller plasma
histamine content was found to be very low, while it was high
in the 6 animals which reacted with bronchoconstriction. In
none of the animals with the greater plasma histamine con-
centration was there a change of the bronchial tone, in spite
of the fact that in some controls a high histamine sensitivity
was revealed.

Figure 10 illustrates one of these experiments (pair 9).
During cross circulation a bronchoconstriction was elicited in
the animal with the low plasma histamine concentration. The
histamine sensitivity of the bronchi was high in both animals.

The objection may be raised that it has not been proved that
the bronchoconstriction in these experiments was caused by
histamine. Yet it is remarkable that during cross circulation
between an animal of a particularly great and another of a
particularly small plasma histamine content a typical hista-
mine effect is elicited in the latter; this occurs only if the
histamine sensitivity is high. In the animal of great plasma
histamine content no bronchoconstriction occurs even if the
histamine sensitivity is high.

These experiments strongly support the conception, that
histamine is present in the circulating plasma in an active state.
There still remains the objection that an extremely labile com-
pound might have been decomposed when the blood passed the
system of cannulae and tubings; this assumption must, however,
be rejected, since bronchoconstriction never occurred in the
animal with the higher concentration of this hypothetical in-
active compound in the plasma.

Chapter 6

Experiments where the histamine content of
plasma is artificially increased

In these experiments we have raised the plasma histamine
concentration by slow intravenous injection of histamine during
a long period and have then tried to find out in what form this
histamine appears in the circulating plasma. Using a special
method we have been able to greatly augment the plasma
histamine concentration in cats and dogs. Dale and Laidlaw
(1919) found, that the intravenous injection of 100 y histamine
per minute to a cat weighing l.s kg (56 y per kg and minute)
caused a severe shock. Me Carrell and Drinker (1941) showed,
that in dogs a histamine quantity of the same order caused a
decrease of the blood pressure of long duration. If, on the
other hand, a smaller amount of histamine (about 1.5 y/kg and
min.) is slowly injected into dogs the blood pressure first falls
and then rises again in spite of continued injection; within
5 — 18 minutes the pressure returns to its original level (Giraud-
Costa and Gayral 1940). In cats and dogs we have been able
to confirm this and we have further made the following obser-
vation. After the original blood pressure has been re-established
the slow injection can be continued with double the amount
of histamine, the blood pressure then showing only a very
transient fall. The dose can be doubled repeatedly, each in-
crease in dose causing only a small or no fall in blood pressure.
By this method it is possible to inject histamine in concen-
trations which would kill an untreated animal.

Many investigators have observed that the histamine sensitivity of
the body can be diminished by previous histamine injections. Dale and

Laidlaw (1911) injected an amount of histamine into a cat which
caused severe symtoms and found that the double amount elicited a
considerably smaller reaction when given shortly aftenvards. Similar
observations have been made by Fuehner (1912), Ochme (1918), Eichler
and Killian (1931), MUller ct al. (1932), A decrease in histamine sensi-
tivity of blood vessels after previous injections of histamine has been
obsen-ed by several investigators (Phemister and Handy 1927, Epstein
1932, V. Euler 1938, Anrep et al. 1939, Hcidcmann 1943). These obser-
vations have been made in acute experiments. Other investigators have
made e.xperiments of a somewhat different type. They have tried to
find out if the histamine sensitivity can be altered by a series of
histamine injections, c. g. daily injections during a period (Striimbcck
1932, Schiff 1938, Farmer 1939, Edholm 19-12, Katzenstcin 1944). In this
connexion it is also of interest that many investigators have tried to
protect experimental animals against anaphylactic shock bj' histamine
pretreatment (Smith 1939, Farmer 1939, Arloing ct al. 1939); histamine
injections have also been used in the treatment of allergic and other
conditions, whore histamine liberation is believed to play a part (Horton
et al. 1936, Horton 1941, Rainey 1943, Butler and Thomas 1945).

Our investigation was carried out on cats under chloralose
and dogs under morpliine-chloralose. In some experiments
artificial respiration xvas used. The blood pressure in the
carotid or femoral artery wa.s registered by a mercury mano-
meter. Histamine biphosphate, dissolved in Tyrode solution,
was slowly injected by a special device into a femoral vein.
The solutions were of relatively high concentration so as to
avoid the injection of great amounts of fluid (O.a — 0.4 ml per
minute was given). Separate injections xvere made through a
cannula, inserted in the other femoral vein. From a cannula
in a femoral artery 5 to 16 ml blood was collected in a centri-
fuging tube, containing heparine. Plasma xvas extracted or
ultrafiltrated and then tested on guinea-pig’s ileum.

The results of these experiments are summarized in tables
18 and 19. These experiments show that the histamine con-
centration of plasma can be raised to very high levels without
a marked fall in blood pressure, provided that the high level
18 established successively. The injected histamine is rapidly
e iminated. It is well known that injected histamine very
mpidly disappears from the blood (Anrep and Barsoum 1935,
Dragstedt and Mead 1935, Mac Intosh 1938, Rose and Browne

TABLE 18,

Injection of histamine

Histamine content of
plasma

Blood pressure,
mm Hg

Duration of
injection,
minutes

Total amount
of histamine
injected y/kg

Largest dose
of histamine
injected f/kg
and rain.

before

injection

y/1

after injection

before

injection

6 ^

0«3 O

ft ®

.5 c r-p
o c c

©ts ^

rcf »

r/1

O .-S *
C4 j: ^ c

100

470

140

170

145

cat

750

150

cat

120

1850

205

150

160

cat

100

150

715

140

210

cat

500

130

160

135

cat

1125

180

140

125

170

cat

100

630

100

145

130

cat

1938, Billings and Maegraith 1938, Code 1939, Hildebrand
1940).

Some of these experiments have been used for the identifi-
cation of histamine in the ultrafiltrates as described in chapter 3.
Our chief aim, however, was to investigate if the histamine,
which can be detected in plasma after the injection, is present
in vivo in a physiologically active state or has been inactivated

TABLE If).

Injection of histamine

Histamine content of
plasma

Blood pressure,
mm Hg

C GC

t g.S
S'c E
O'"

s'U

S -S
^.2 «

0»*-**c*

O O

05 C ^

S-c c
<U to ^ VM
o g
h3 0.5

£.2_

o ft»

■°'b

after injection

o o

5 u

U ^ ^

O CD 2 ^
r:^ ^ ^ t:

o« c:.c

O ^ ^

W-®-? C
a £■••■
'C o E
cs «

y/’l

*>« ^

C3 — C

® n." E
E “.c

1940

250

1800

130

cat

750

120

200

160

cat

300

3825

200

2000

290

200

210

IGO

135

125

110

100

cat

230

2275

t «
!«

185

200

100

cat

320

7150

120

1000

—

cat

315

20G50

5G0

—

■

240

180

cat

285

18400

800

■

140

105

dog

975

300

140

110

dog

220

5900

200

4300

IGOO

0.5

—

110

by combination e. g. with some plasma component. The fol-
lowing three observations indicate that histamine in these
experiments is present in plasma in a physiologically active
state:

1) the plasma ultrafiltrate from blood, withdrawn after slow
injection of histamine is found to contain the same amount of
histamine as the plasma extract when tested on the gut. Table
20, in which the animals are numbered as in tables 18 and 19
gives the histamine concentrations of ultrafiltrates and extracts.

TABLE 20.

Histamine content of plasma, f/l

before injection

after injection

extraction

ultrafiltration

extraction

ultrafiltration

150

100

1800

1850

120

130

—

2000

420

400

200

1000

—

1400

1550

—

7800

300

320

4300

4500

2) the more the plasma histamine concentration is increased,
the smaller the fall in blood pressure becomes, caused by a dose
of histamine intravenously injected. This fact is illustrated by
figure 11. In this experiment, before the slow histamine injec-
tion, the plasma contained 30 y histamine per litre, and 0.3 y
histamine, given intravenously, caused a fall in blood pressure
of about 80 mm Hg. After the histamine content of the plasma
had been raised to 140 y per litre, 0.3 y histamine caused a
fall of only 20 — 30 mm. When the histamine concentration had
decreased to 95 y per litre, the fall in blood pressure was about
50 mm, and when the histamine concentration had returned to
nearly the original level the histamine sensitivity was almost
restored.

i. ''>”' sC *”" '“‘'f i° c^ “ ■hi

iirsv nr^r n‘ “b^- Sbh;

““*• f rSit » ®”“‘® "'o »* ® ® ”'“”‘Se’6 s«‘' ’■> ®a

of to 4 ) immediately

litre. ) . , vet it seems

„V>.vc4fflere»te.^ana.-:;V-»-^
T^eseltaMasa tt>eWata®« 5,^ state.

«.so.aWe to oon^ ^ p„jsiotosSteaUy

peseat ta plasma .„

GaJ 4 »» fe "isb'y ^“jrimoatB “ i„ htetaaim”.

isolated piece S^^^vutioa. Our expe oTgato«\ y^istamme-

a atrong lusta® to dcao ^ p.etreatmeut ^^der

tSected or cSto^® - !ftst''Sd ^,3wity is

9.) it tte ''“““'J^Sstamtoo lBi®“.”"Vistaiatos '"

s;’,r.:-s - “— "

Fig. 12. Registrations from above: bronchial tone (according to Konzett
and RSssler), jejunal motility (registered with a water manometer, con-
nected with a balloon, inserted in the jejunum through an incision in
the abdominal wall) and blood pressure in the femoral artery of the
recipient cat B; blood pressure in the carotid artery of the donor cat A;
time in minutes; signal marks. 1: transfusion of blood. 2*. 100 y histamine
to cat B. 3: 50 y histamine to cat B.

the recipient. Four such experiments were made giving similar
results. Figure 12 illustrates such an experiment.

In a cat A weighing 3.3 kg (cat nr 15 of table 19) histamine was
slowly injected in increasing doses (1.2 — 800 y per kg and minute)
for nearly five hours. On another cat B (1.7 kg) the bronchial tone, the
jejunal motility and the blood pressure were registered. A cannula,
placed in the central stump of the femoral artery of cat A was connected
by means of a piece of rubber tubing with another cannula inserted in
the femoral vein of cat B. The connecting system was filled with Tyrode
solution and the rubber tubing was occluded by a clip. Heparine was
injected into both animals. Three minutes before 1 in figure 12 the
injection of histamine into cat A was interrupted. At 1 the clip was
opened for 25 seconds and the blood allowed to flow from cat A to
cat B. As can be seen from the figure the transfer of blood from the

!)7

donor caused bronchial constriction, contraction of tlie put and fall in
blood pressure in the recepient. The cannula was then removed from
the femoral vein of cat B. The slow injection of histamine to cat A
was then repeated for a few minutes. Three minutes later the clip was
again removed and blood allowed to flow through the transfusion system
into a raensure. 15 ml of blood was collected in 2o seconds. At 2, 100 */
histamine dissolved in 15 ml warm Tyrodc solution was injected intra-
venously for 25 seconds to c.at B. At ,% 50 r histamine was given in
the same way. The blood sample was used for detenuination of the
hematocrit value of the blood and for estimation of the histamine content
of the plasma on guinea-pig’s gut. It was found that the blood contained
62 per cent of plasma and that the plasma contained 7S00 y histamine
per litre. Presuming that 15 ml blood containing 7800 */ histamine per
litre was transfused, cat B should then have received 70 — ^75 y histamine.
It can be seen in figure 12 that 50 — 100 ;• histamine, injected in a
similar way, has an effect of the same order as the transfused blood.
It is obvious that the figures obtained in this way are not very exact;
itm.ay be mentioned that the quantitative relations might be complicated
e.g. by the fact that the blood of the donor probably had a relatively
high concentration of adrenaline which might have interfered with the
assay. Nevertheless the experiment indicates that most of the histamine,
which was delected in the plasma of the donor (7600 y per litre), was
present in a physiologic.ally active state.

These series of experiments show, that histamine can be
present in plasma of cat and dog in a concentration, which
considerably exceeds the normal level without serious depression
of blood pressure. In these experiments the histamine was not,
at least not to a large extent, bound as an inactive compound;
histamine occurred in plasma in a pliysiologically active form.

Chapter 7

Discussion

The discovery, that physiologically highly active substances
such as acetylcholine and histamine are present in the body
in high concentrations, introduced an interesting physiological
problem. Best, Dale, Dudley and Thorpe (1927) in their paper
on »the nature of the vasodilator constituents of certain tissue
extracts® suggested two explanations. The active substance
may exist »only potentially, in form of some inactive precursor®
or it may be »present as such in the cell interior, being pre-
vented from leaving it so long as the cell membrane is physio-
logically intact®. Recently several workers have dealt with the
problem of »free» and abound® histamine and acetylcholine.
In this connexion experiments by Abdon et al. (1939, 1944,
1945) are of special interest from which they conclude, that
in the body there exists a labile complex compound, from which
acetylcholine is very easily liberated but which does not itself
exert the effects typical to acetylcholine. Many observations
suggest that histamine is present in the cells in an inactive
form, but there is so far no experimental evidence of the
existence of a histamine precursor.

With the discovery, that histamine is a normal constituent
of blood the question arose whether this histamine is physio-
logically active or inactive. Gaddum (1936) called attention
to this problem. The problem was experimentally dealt with
by Tarras-Wahlberg (1936).

In our experiments w^e have tried- to find out if histamine
is present in plasma in a physiologically active or inactive state.
As is well-known histamine, present within intact blood cells,

does not exert any effects (Anrep et al. 1936, Barsoum and
Smirk 1936). At the beginning of our experiments it seemed
reasonable to assume that the plasma histamine was present
in a »bound» state; considering the high physiological activity
of histamine and also the histaminolytic power of tissues and
plasma it is difficult to conceive that histamine can exist in
plasma in -a free state.

Our experiments have given the following results. In an
ultrafiltrate of a plasma sample histamine is present in a phy-
siologically active state. This is the case even if the ultrafiltrate
is obtained by ultrafiltration in vivo when the blood is handled
with the utmost care. By ultrafiltration approximately the same
amount of histamine is recovered from a plasma sample as by
chemical extraction. From these experiments it can he concluded
that histamine is not present in plasma as a stable, inactive
compound; that histamine should have been produced from
histidine at the preparation of the plasma for the assay is out
of the question. Kaiser (1939) is of the opinion, that the plasma
histamine is adsorbed to protein and physiologically inactive;
he believes that histamine, added to plasma, is rapidly inac-
tivated by adsorption. The ability of plasma proteins to combine
with other substances is well established and has been referred
to as its »Vehikelfunktion». If for instance certain digitalis
glucosides, such as digilanid A, are mixed with blood serum,
they are partly adsorbed by albumen and their pharmacological
activity is weakened (Suter 1944). Trypanblue, intravenously
injected, is bound to proteins (Hechter 1943). A portion of the
calcium, normally present in plasma occurs as a protein com-
pound; if calcium is injected intravenously, part of 'this calcium
combines with protein (Rona and Takahashi 1910, Cushny 1920,
D’Silva 1934, Cameron and Moorhouse 1937). Histamine can,
as a result of our experiments, not be present in plasma bound
in a similar way. The three substances referred to above can
not be detached from the proteins by ultrafiltration or dialysis,
whereas the same amount of histamine by ultrafiltration can be
recovered from plasma as by chemical extraction; this is appli-
cable to the normally present as well as to injected histamine.
There is also another difference. The fraction of the plasma

calcium, which is bound to protein, is not contained in the
cerebrospinal fluid. Our experiments show, that histamine is
present in the aqueous humour in about the same concentration
as in plasma. The resemblance between aqueous and cerebro-
spinal fluid as to mode of formation and composition has often
been stressed.

Our experiments with ultrafiltration in vitro and in vivo and
with aqueous humour indicate, that if histamine is normally
present in plasma as a physiologically inactive compound, this
compound must be extremely labile. The cross circulation
experiments strongly support the view that histamine exists in
circulating plasma in a physiologically active form. It is also
demonstrated that intravenously injected histamine can circulate
for some time in an active state in the plasma.

We have considered it necessary to make two kinds of control
experiments. Firstly we have tried to ascertain that the active
substance found in plasma samples really is histamine. In a
physico-chemical respect the substance corresponds to hista-
mine in being ultrafiltrable and not destroyed by Code’s ex-
traction procedure. As to the biological properties we have
found a close agreement. The substance contracts the isolated
guinea-pig’s gut; this effect, like that of histamine, is antago-
nized by thyraoxyethyldiethylamine or theamine. It depresses
the blood pressure of the atropinized cat when injected intra-
venously; on close arterial injection in the adrenals, the agent,
like histamine, causes a rise in blood pressure. The ultrafiltrate
does not give a fall in blood pressure in a cat which has
previously been treated with histamine. It constricts isolated
guinea-pig’s bronchi. The identity between the agent of the
ultrafiltrate and histamine is also suggested by quantitative
relations: both have the same concentration-action curve on
isolated guinea-pig’s gut; w'hen matching the ultrafiltrate against
histamine equal results are obtained on different test objects.
The ultrafiltrate has no significant effect on the rabbit’s blood
pressure and other test objects not affected by histamine in
moderate concentrations. The active agent is identical with
histamine in every respect investigated; no substance other than
histamine is known, which has all the properties described here.

We have further tried to ascertain that histamine really is a
normal constituent of the circulating plasma and that the hista-
mine found in a plasma sample, in vitro, has not been liberated
from the corpuscles when securing the sample. The experiments
on rabbit’s blood show, that clotting or rough handling of the
blood can cause a considerable increase of the plasma histamine
concentration; a carefull treatment of the blood, on the other
hand, and especially ultrafiltration in vivo gives samples with
low histamine concentration. Yet many of our observations
support the assumption that histamine is normally present in
the circulating plasma; in guinea-pigs, rats and dogs the plasma
histamine concentration is maintained at a fairly constant level
during long periods; histamine is present in in-vivo-ultrafiltrates
from dogs and rabbits and in aqueous humour from guinea-pigs,
rats, rabbits, cats and dogs; in the species investigated the
histamine concentration of plasma in vitro, aqueous and in-vivo-
ultrafiltrate is of the same order. The plasma samples from
rabbits are quite an exception, but there is reason to believe
that also in rabbits histamine is present in the circulating
plasma in the same concentration as in aqueous and in-vivo-
ultrafiltrate. The cross circulation experiments in guinea-pigs
strongly suggest that histamine is really contained in the cir-
culating plasma.

Summarizing all our experimental evidence it can be con-
cluded that histamine is really present in the circulating plasma
in a physiologically active state.

Our experiments may contribute to the understanding of the
puzzling fact, that this highly active substance can be present
in plasma. From the experiments in chapter 6 it is obvious
that the circulatory system has a considerable adaptability to
high plasma concentrations of physiologically active histamine,
provided that the concentration is increased slowly. In these
experiments the histamine was injected in gradually increasing
doses, but apart from this no measures were taken which could
justify the assumption that an adaptation to histamine should
be possible only under these special experimental conditions.
On the contrary, it must be remembered that in experiments of
this type anaesthesia and loss of blood (due to withdrawal of

blood samples) greatly decrease the tolerance to histamine
(Dale 1920). We have not investigated the mechanisms which
maintain the blood pressure at a high level in these experiments.
Eeflex mechanisms should be involved. Apart from that, it
seems likely that the effector cells can adapt themselves to
high histamine concentrations just as the cells of the isolated
gut, bathed in a strong histamine solution. Besides, antagonistic
substances may counteract the effect of histamine present in
plasma. Histamine causes an output of adrenaline from the
suprarenals; a physiological antagonism between these two
substances has been discussed (Dale and Richards 1918, Burn
and Dale 1926). Other substances may also play a part; Acker-
mann and Wasmuth (1939) suggest, that the antagonism between
histamine and arginine or histidine may be of physiological
significance.

If the conception that active histamine is present in plasma
is accepted, another problem arises, which has in our introduc-
tion also been referred to: How is it possible that histamine
can exist in plasma, where it is subjected to a histaminolytic
activity? It should, however, be stressed that our experiments
only indicate that histamine is present in an active state; they
do not prove that it occurs in a free form. The possibility still
remains that the plasma histamine is active, but bound as a
compound where it is protected against the histaminolytic
activity. Dale (1932) suggested that acetylcholine might exist
in an active complex compound, in which it can not be attacked
by the acetylcholine esterase. That histamine could exist in a
bound state and yet be physiologically active has also been
suggested by Dale (1937 — ^38); in such a hypothetical compound
it seems necessary that two groups of the histamine molecule
should be free: the NHj-group, to which the specific activity is
due, and the NH-group, by which the histamine is anchored to
the tissues (Rocha e Silva 1944). It may be added that Sachs
et al. (1932) have observed, that histamine can exist in an
adsorbed state and yet have a depressant and secretory activity.

The problem under discussion can also be explained in another
way. It may be possible that histamine is permanently destroyed
and that there is at the same time a continued inflow of hista-

mine into the blood. This histamine may originate from metabolic
processes of the tissues (see e. g. Burn and Dale 1926, Dale
1927, 1929, 1933, Lewis 1927, Gotzl and Dragstedt 1940), or it
may be derived from the food, possibly produced by bakteria
in the gut (Mellanby and Twort 1912, Abel and Kubota 1919,
Koessler and Hanke 1924). If this be true there must exist some
regulating mechanism which effectively balances inflow and
elimination of histamine. The fact that the histamine concen-
tration in one and the same animal is maintained fairly constant
for a long period and that, after histamine injections, the con-
centration rapidly returns to the original level, indicates that
such a regulating mechanism really exists.

The finding that histamine is present in plasma in an active
form gives rise to the question if this histamine has any physio-
logical function. On this point nothing is known with certainty,
and although such functions may come within many domains
it is tempting to refer to the hypothesis of a ^balanced chemical
control of capillary tone» in which histamine is believed to play
a part (Dale and Richards 1918, Burn and Dale 1926).

Summary

Experiments have been carried out on guinea-pigs, rats,
rabbit-s, cats and dogs.

1. Ultrafiltrates of plasma from these animals contain an
agent -which contracts the isolated, atropinized guinea-pig’s
ileum; this effect is antagonized by thymoxyethylediethylamine
or theamine. The activities of the ultrafiltrate and of histamine
follow the same concentration-action curve. The plasma ultra-
filtrate depresses the atropinized cat’s blood pressure when
injected intravenously; on close arterial injection in the adrenals
it raises the blood pressure. The bronchi of the perfused guinea-
pig’s lung are constricted by the ultrafiltrate. It is concluded,
that ultrafiltrates of plasma contain an agent which is identical
with histamine.

2. The histamine concentration of plasma from guinea-pigs,
rats and dogs is maintained at a fairly constant level during
long periods. This indicates that histamine is a normal con-
stituent of plasma and that the histamine present in plasma
samples in vitro has not been liberated from the blood cor-
puscles. This conception is further supported by the observations
mentioned below.

3. A plasma ultrafiltrate contains histamine in about the
same concentration as a corresponding sample, extracted che-
mically. The histamine content of aqueous humour and of in-
vivo-ultrafiltrates is approximately the same as that of plasma.
From these findings it is concluded that histamine exists in
plasma in a physiologically active form or as an inactive
compound, from which histamine is extremely easily liberated.
Cross circulation experiments on guinea-pigs indicate that hista-
mine is present in an active form in the circulating plasma.

4. B 3 " slow intravenous injection of histamine at increasing
rates the histamine concentration of the plasma can be con-
siderabl}’- raised in cats without any serious depression of the
blood pressure. It is concluded that histamine, injected in this
waj% is carried by the plasma in an active form.

Acknowledgement

I wish to express my deep gratitude to professor Georg
Kahlson for not only his great help and encouragement hut
also for giving me such excellent working facilities. My best
thanks are also due to miss Greta Emanuelsson for valuable
technical assistance. To my wife, med. kand. Kerstin Emmelin,
who assisted me so much both with the experiments and the
translation into English, I express my deep appreciation and
thanks.

The work has been supported by grants from »Doktor Torsten
Anders Amund Amundsons Fond», ^Stiftelsen Therese och Johan
Anderssons minne» and »Fil. Doktor P. Hilkanssons stiftelse».

References

Abdok, N.-O. and S. 0. Hammar.sk.jOlu, K. Fysiograf. Sallsk. Fcirh. 1939.
9. Hr 15.

— Acta Ph 3 *siol. Scand. 1944. 8. 75;

Abdok, N.-O. and K. Ljungdaiii.-Ostberg, Ibid. 1944. 8. 103.

Abdon, N,-0. and T. Bj.wke, Acta Pharmacol. Toxicol. 1945. 1. 1.

Abel, J. J. and S. Kubota, J. Pharin. exp. Ther. 1919. 13. 243.

Ackermans, D. and W. Wasmuth, Z. physiol. Chem. 1939. 25. 259. 28.
.\HLM.ARK, A., Acta Physiol. Scand. 1944. 9. Suppl. 28.

Ahlmark, A. and T. G. Kornerup, Nord. Med. 1939. 2. 1803.

Alam, M., G. V. Akrep, G. S. Barsou.m, M. Talaat and E. Wieninger,
J. Physiol. 1939. 95. 148.

Akrep, G. V., N. S. Ayadi, G. S. Barsoum, J. R. Smith and M. Talaat,
Ibid. 1944. 103. 155.

Anhep, G. V. and G. S. Barsoum, Ibid. 1935. 85. 409.

— Ibid. 1935. 85. 36P.

Anrep, G. V.,.G. S. Barsoum and M. Tala.vt, Ibid. 1936. 86. 431.

Akrep, 6. V., G. S. Barsoum, M. Tala.at and E. Wieninger, Ibid. 1939.
95. 476.

— Ibid. 1939. 36. 130.

Arloing, F., L. Thevekot and J. Viai.lier, C. E. Soc. Biol. Paris 1939.
130. 659.

Barger, G. and H. H. Dale, J. Physiol. 1911. 41. 499.

Barsou-m, G. S. and J. H, Gaddum, ibid. 1935. 85. 1.

Barsoum, G. S. and F. H. Smirk, Clin. Science 1936. 2. 337.

Bartosch, R., W. Feldberg and E. Nagel, Arch. ges. .Physiol. 1932. 230.
129.

Best, C. H., H. H. Dale, H. W. Dudley and W. V. Thorpe, J. Phvsibl.
1927. 62. 397.

Best, C. H. and E. W. McHenry, Amer. .1. Physiol. 1930. 93. 633.

— Physiol. Rev. 1931. II. 371.

Billings, F. T. and B. G. Maegraith, Quart. J. exp. Physiol. 1938. 27. 249.
Bovet, D. and A. M. Staub, C. R. Soc. Biol. Paris 1937. 124. 547; 125. 818.
Buchthal, F. and G. Kahlson, Acta Physiol. Scand. 1944. 8. 325.

Buell, M. V. and M. E. Perkin.s, J. biol. Chem. 1928. 76. 95.

Burn', J, H. and H. H. Dale, J. Physiol. 192C. 61. 185.

Butler, S. and W. A. Thomas, J. Amer. Med. Assoc. 1945. 128. 173.
Ca.meron, a. T. and V. H. K. Moorhouse, J. Physiol. 1937. 91. 90.

Chute. A. L. and E. T. Waters, Amer. J. Physiol. 1941. 132. 552.

Code, C. F., J. Physiol. 1937. 89. 257.

— Ibid. 1937. 88. lOP.

— Ibid. 1937. SO. 485.

-- Ibid. 1937. 90 349.

— Amer. J. Physiol, 1939. 127. 78.

Code, C. F., C. L. Evans and R. A. Gregory, J. Physiol. 1933. 92. 344.
Code, C. F. and H. R. Inu, Ibid. 1937. 90. 501.

Code. C. F. and A. D. MacDonald, Lancet 1937. 233. 730.

Cubhny, a. R., j. Physiol. 1920. 53. 391..

Dale, H. H., Brit. J. exp. Path. 1920. 1. 103.

— Lancet 1929. 216. 1179, 1233, 1285.

— Verb, deutsch. Ges. inn. Med. 1932. S. 17.

— Bull. Johns Hopkins Hospital 1933. .53. 297.

— 1936, in Gefiisserweiternde Stoffe der Gewebe, Leipzig 1936. Bv
J. H. Gaddum.

— Verhandl. deutsch. Ges. Kreislaufforsch. 1937-- 38. 13.

Dale, H. H. and P. P. Laidlaw, J. Physiol. 1910. 47. 318.

— Ibid. 1911. 43. 182.

— Ibid. 1919. 52. 355.

Dale, H. H. and A. N. Richard.<. Ibid. 1918. 52. 110.

Davson, H., Ibid. 1939. 96. 194 ’

Davson, H. and J. P. Quilliam, Ibid. 1940. 98. 141.

Draostedt, C. A., Quart. Bull. 1940. 14. 283.

— J. Allergy 1945. 16. 69.

Duagstedt, C. A. and F. B. Mead, Proc. Soe. exp. Biol. N. Y. 1935. 32.
1435.

D’Silva, j. L.. J. Physiol. 1934. 80. 6P.

Duke-Elder, S. and H. Davson, Brit. J. Ophtalm. 1943v 27. 431.

Edholm, 0. G., J. Physiol. 1942. 101. 262.

Eichler, 0. and H. Killian, Arch. exp. Path. Pharmakol. 1931. 159. 606.
Emmelin, N. G., Acta Physiol. Scand. 1943. 5. 372.

~ Ibid 1945. 9. 378.

Emmelin, N. G., G. S. Kahlson and K. Lindstrom, Ibid. 1941. 3. 39.
Emmelin, N. G., G, S. Kahlson and F. Wicksell, Ibid. 1941. 2. 123.
Emmelin, N. G. and E. Palm, Acta Ophtalm. 1944. 22. 117.

Epstein, D., J. Physiol. 1932. 76. 347.

Euler, U. S. v., Ibid. 1938. 93. 129.

Farmer, L., J. Immunology 1839. 36. 37.

Feldrerg, W., E. Flatow and E. Schilf, Arch. exp. Path. Pharmakol.
1929. 140. 129.

Feldbkiw, W., Extrait du volume jubilaire publie en I’honneur du Pro-
fesseiir J. Demoor. 1937. S. 181.

Feldberg, W. and W. J. O’Connor, J. Physiol. 1937. 90 . 288.

Feldbeur, W. and E. Schilf, Histamin. Berlin 1980.

Fleiscii, a., Arch. ges. Physiol. 1937. 239. 345.

Freijxd, H., Arch. exp. Path. Pharmakol. 1920. 86 . 266. 88 . 39.

— Ibid. 1936. 180 . 189.

Friedexwald, j. S., Brit. J. Ophtalm. 1944. 28 . 503.

Fueii.n-er, H., Munch, med. Wsclir. 1912. . 59 . 852.

Gaddem, j. H., Proc. Royal. 8oc. Med. 1936. 29 . 1373.

— Gefasserweiternde Stoffe der Gewebe. Leipzig 1936.

Gaddix j. H., C. S. Jang and H. Kwiatkowski, J. Physiol. 1933. 96 . 104.
G.ajdos, a., Presse Med. 1938. 46 . 509.

Gasser, H. S., J. Pharmacol. 1926. 27 . 395.

Giraud-Costa and Gayral, C. r. Soc. Biol. Paris 1940. 134 . 88.

Gotzl, F. R. and C. A. Draostedt, Proc. Soc. exp. Biol. Med. 1940. 45 .
688 .

GrOnwael, a. and B. Inoelman, Acta Physiol. Scand. 1945. 9 . 1,
Gl'ccekiieem, M., Die biogenen Amine. Basel 1940.

Guggenheim, M. and W. Lofklkr, Biocliem. Z. 1916. 72 . 303,

Guttentag, 0. E., Arch. exp. Path. Pharmakol. 1931. 162 . 727.

Harres, K. E., Heart 1927. 14 . 161.

Hechter, 0., Endocrinology 1943. 32 . 135.
llEiDEMANN, H.. Nord. Med. 1943. 21 . 873.

Hiedebrani), F., Klin. Wschr. 1940. 12 . 1.

Hodgson, T. H., J. Physiol. 1938. . 94 . 118.

Holtz, P., Klin. Wsclir. 1937. 45 . 1561.

Horton. B. T., J. Amer. Med. Assoc. 1941. 116 . 377 .

Horton, B. T., G. E. Brown and G. M. Roth. Ibid. 1936. 107 . 1263.

Kaiser, P., Schweiz. Z. allg. Path, und Bakt. 1989. 2 . 3.

Katzenstein, R,, Yale J. Biol. Med. 1944. 16 . 325. Quoted from Brit.

Chem. Physiol. Abstr. 1944. 16 . 325.

Ke.vdai.l, a. I. J.. Infect. Dis. 1927. 40 . 689. Quoted from Gaddum 1936.
Koe.ssler, K. K. and JI. T. Hanke, J. Biol. Chem. 1924. 59 . 803, 889.
Konzett, H. and R. Ros.sler, Arch. exp. Path. Pharmakol. 1940. 195 . 71.
KbPi'ER, A., Ergebn. Physiol. 1930. 30 . 153.

Kwiatkowski, H., .1. Physiol. 1941. 100 . 147.

— Ibid. 1943. 102 . 32.

Leuis, F., The blood vessels of the human skin and their responses.
London 1927.

Mac Gregor, R. G. and S. Peat, J. Physiol. 1933. 77 . 312.

Mac Gregor, R. G. and W. V. Thorpe, Biochem. J. 1933. 27 . 1394.

Mac Intosh, F. C., Quart. J. Exp. Physiol. 1938. 28 . 87.

Marcou, L, e. Athanasiu-Verou, D. Chiriceanu, G. Cosma, N. Gikuold and
C. C. Parhon, Presse Med. 1938. 46 . 371.

Me Cahell, J. D. and C. K. Drinker, Amer. ,1. Physiol. 1941. 133. 64.
Mo Dowale, R. J. S. and J. Me Whan, J, Physiol. 1936. S7. 91 P.
Meleanby, E. and F. W. Twort, J. Physiol. 1912. 45. 53.

Minard, D., Amer. J. Physiol. 1941. 132. 327.

Menard, D. and S. R. Rosenth.al, J. exp. Med. 1939. 70. 415.

Moeeer, E. a., H. Saeomon and G. Zuelzer, Arch. exp. Path. Pharmakol.
1932. 164. 441.

Newton, W. H., J. Physiol. 1932. 75. 288.

O’Connor, J. M.. Arch. exp. Path. Pharmakol. 1912. 67. 195.

Oehme, C., Ibid. 1913. 72. 76.

Phemister, D. B. and J. Handy, J. Physiol. 1927. 64. 155.

PoPiEESKi, L., Arch. ges. Physiol. 1920. 178. 237.

Rainey, J. J.. J. Amer. Med. Assoc. 1943. 122. 850.

Reiiberg, P. B., Acta Physiol. Scand. 1943. .5. 305.

Rigeer, R., Handbueh exp. Pharmakol. 1938. 7. 63.

Robertson, J. D. and P. C. Wieeiams, J. Physiol. 1939. 95. 139.

Rocha e Sieva, M., J. Pharm. exp. Ther. 1943. 77. 198.

— Ibid. 1944. 80. 399.

— J. Allergy 1944. 15. 399.

Rocha e Silva, M. and S. 0. Andrade, J. biol. Chem. 1943. 149. 9.
Rona, P. and D. Takahashi, Biochem. Z. 1910—11. 31. 336.

Ro.se, B., Science 1940. 92. 454.

Rose, E. and J. S. L. Browne, Amer. J. Physiol. 1988. 124. 412.

— Proc. Soc. exp. Biol. Med. 1940. 44. 182.

Sacks, J., A. C. Ivy, J. P. Burgess and J. E. Vandoe.ui, Amer. J. Physiol.
1932. 101. 331.

ScHirr, L., Arch. int. Med. 1938. 61. 774.

Schwartz, A., C. R. Soc. Biol. Paris 1936. 123. 1181.

Sibul, I., Arch. ges. Physiol. 1935. 235. 742.

Simon, A., Arch. exp. Path. Pharmakol. 1937. 187. 678.

— Ibid. 1938. 190. 273.

Simon, A. and E. Komeos, Ibid. 1939. 192. 701.

Smith, E. A., Amer. J. Physiol. 1939. 126. 631.

Steggerda, F. R., H. E. Essex and F. G. Mann, Ibid. 1935. 112. 70.
StrOmbeck, j. P., Skand. Arch. Physiol. 1932. 65. 92.

SuTER, E., Helvet. Physiol. Pharmacol. Acta 1944. 2. 211.

Szczygieeski, j.. Arch. exp. Path. Pharmakol. 1932. 166. 319.
Tarras-Wahlberg, B.. Skand. Arch. Physiol. 1936. 73. Suppl. 1.
Trethewie, E. R., Austral. J. exp. Biol. 1938. 16. 225.

Tsai. C.. J. J. McBride and M. B. Zucker, Proc. Soc. exp. Biol. Med.
1944. 55. 283.

Vogt, M.. J. Physiol. 1943. 102. 170.

Wole, H. J. and H.-A. Heinsen, Arch. exp. Path. Pharmakol. 1935. 179. 15.
Zadina, R.. C. R. Soc. Biol. Paris 1939. 132. 31.

rati.. 1 ? 3 . 1- ItepW'*

ActaW-''’'-

ACTA PHl'^SIOLOGICA SCANDINAVICA

VOL. 11. SUPPLEMENTUM XXXV.

From the Physiological Department, Karolinska Institutet.

Stockholm

A STUDY OF THE RESPIRATORY
REFLEXES ELICITED
FROM THE AORTIC AND
CAROTID BODIES

BO E. GERNANDT

Stockholm 1946

Contents

Preface

Introduction '

I. The existence of extra-cnrotid clietnoreceplors , . 9

II. Anatomical puney

The aortic and pnlinonary pnrapanfzlia I.'I

III. Aortic body

1. Technique and procedure

2 . Localization

3. Nerve supply

4. Blood RUpj)ly

IV. Tlie effect of hypoxemia and hypercapnia on the

chenioreccptors in the aortic body 33

1. Hypoxemia 35

2. Hypercapnia 39

Tile effect of various drugs on the chenioreceplors

in the aortic body 41

1. Lobelinc 41

2. Piperidine 4:1

3. Cyanide 4.5

4. Acelylcboline 4(1

Selective elimination of the chemorcccptors in the
carotid and aortic hodics 48

1. Technique and procedure 50

2. Carotid body 50

3. Aortic body 55

The distribution of the effect of specific stimuli

between the carotid and aortic hodics 56

L Technique and procedure 56

Poge

2. Hypoxemia 57

a) The respiration 57

b) The blood pressure 58

3. Lobeline 59

a) The respiration 59

b) The blood pressure 60

4. Hypercapnia 61

a) The respiration 61

b) The blood pressure 61

Vni. The reflex effect on the respiration of intrasinusal

pressure changes 62

IX. Adrenaline apnoea 66

1. The effect of intravenous injection of adrenaline

on the activity in the chemoceptive fibres in the
depressor nerve 68

2. The effect on the respiration of intravenous

adrenaline injection after elimination of the
chemoreceptors 69

X. Tonic chemoreflex respiratory stimulation .... 70

1. The effect on the respiration 72

2. The effect on the blood pressure 73

XL Summary 76

References 78

PREFACE

The present work has been carried out at tlie Physiological
Department, Karolinska Institutet, where for many years par-
ticular interest has been devoted to tlie reflexogenie effect under
various conditions upon the respiration and circulation elicited
from the carotid sinus region. It is in this connection a great
pleasure for me to express my warm thanks to the head of this
department. Professor Ulf von Euler, who aroused my interest
in the first place for physiological research. In the course of the
work he has ahvays discussed with me tlie problems arising ^vith
great willingness and interest, and here his long experience in
this field has been of great help to me.

To Professor Yngve Zotterman, with whom during the past
years I have had the great privilege of cooperating, and from
whom during this period I have learned the electro-physiological
technique, it is a sincere pleasure for me to express my deep
gratitude for all valuable advice as well as criticism and encour-
agement in connection ivith this ivork. I am also much indebted
to him for placing his laboratory and apparatus at my disposal.

For tlie assistance Avhicli my wife has always so readily given
me, and which has been a great time-saving help, I will here
express my warm thanks.

To Mr Donald Burton, avIio has performed the translation
into English, I am also much indebted.

The investigation has been carried out Avith the help of econo-
mic support from the Theresa and Johan Andersson Memorial
Foundation and tlie Christian Loven Foundation, for Avhich help
I Avish to express my thanks.

Stockholm, March 1946.

Bo E. G E R N A N D T

INTRODUCTION

Since the discovery of the chemoreceptors in the carotid sinus
region hy Heymans, Boucicaert and Dautrebande (1930) these
have been made the object of investigation by a very large number
of physiologists, so that tlieir function has now been made clear
in many essential respects. Even several years earlier, however,
J. F. Hetmans and C. Heymans (1927) had shown the existence
and importance of such receptors in the region around the
beginning of the aorta. As these chemoreceptors, as compared
with those localized to the carotid sinus, are of more subordinate
importance and do not, moreover, from the technical point of
view, offer such favourable experimental conditions, they have
been less sought after for special physiological studies. In experi-
ments on the reflex effect from the chemoreceptors on respiration,
blood pressure, pulse frequency, width of vessels, adrenaline
secretion etc., most writers, taking as their point of departure the
fundamental investigations of Heymans and Heymans, have only
had the chemoreceptors of the aorta in mind, but tlie main
interest has been directed to the receptors in the carotid bodies.
Only CoMUOE (1939) has made the chemoreceptors of the aorta
the object of renewed direct investigation.

Long before the attention of the physiologists had been directed
to the problems in this particular field and before they had been
able to provide an explanation for them, the morphologists knew
of the occurrence of so-called paraganglionic formations in these
regions.

The intention behind the present investigation has been •vvitli
the aid chiefly of electro-physiological recording of the impulses
in tlie nerves from the chemoreceptors of the aorta to study more
closely the effect of chemical changes in the blood on the inflow
of afferent impulses. Also the effect of drugs having a specific
stimulating effect on the chemoreceptors has been studied.

By means of a method of selectively eliminating the chemical
reflexes from the aorta and sinus regions, v/hile the pressor
reflexes were left intact, a different approach to problems which
have been the subject of diverging opinions was made possible.
This procedure makes it possible to leave the buffer nerves
regulating the blood pressure, the sinus and depressor nerves,
intact. Previously, in the study of tlie effect reflexly elicited from
the chemoreceptors, a selective elimination of the chemoreceptors
of the aorta has not, apart from the denervation procedure, been
practised. On the other hand, it has been possible mechanically
to destroy the chemoreceptors in the carotid sinus while the in
other respects intact sinus nerve still conducted pressor impulses.
The beyond all comparison most common method, however, has
been to sever the sinus and depressor nerves (or the vago-
depressors), in order to exclude tlie effect from the chemo-
receptors. In consequence of the fact tliat the buffer nerves had
then been eliminated, the blood pressure has shown such changes
that the effects investigated have been influenced, and a pure
picture of the actual mechanism has thus not been obtained.

After selective elimination in order of the chemoreceptors of
the aorta and sinus regions, it has been possible to investigate the
distribution of the reflexly stimulating effect over the respective
chemoreceptors on specific stimuli. The eventual inhibiting
effect on the respiration released from the baroceptors has also
been made the object of special interest.

1. The existence of extra-carotid
chem oreceptors

In a work published in 1927 J. F. Heymans and C. Heymans
■were the first to sho'W tliat the respiration could be affected
reflexly from the thoracic region, in which connection chemical
stimuli served as tlie eliciting factor. The effect also appeared,
as the authors were able to show, if one only perfused the heart
and tlie nearest part of the aorta of dog, the head being isolated
from the body mth tlie exception of both vago-depressors. An
increase of the respiratory movements of the head was brought
about when tlie content of carbon dioxide in tlie blood was
increased or if its oxygen-content was reduced. The only path
the respiration-stimulating impulses could take was over the
vago-depressors, and it was thus shoivn tliat chemical changes
in the blood could elicit respiratory reactions by means of
reflexes elicited from structures outside the respiratory centre.
For the chemical regulation of the respiration was formerly
considered to take place exclusively over the respiratory centre
in tlie medulla oblongata, but through the investigations of the
above authors these problems appeared in a totally different
light, interest in these questions was immensely stimulated, and
in the course of the years an extremely copious literature on the
subject has seen the light.

Heyaians, Boucicaeut and Dautrebande (1930, 1931) were
aftenvards able to show that analogous conditions existed also in
the carotid sinuses. In these experiments on dog, tlie in respect
of the circulation isolated sinus region ivas perfused with the aid
of another animal or by means of a pump, whereby all central
effects could be excluded. That it was possible to provoke respi-
ratory reactions from these regions through endosinusal alterna-
tions of pressure had already been shown by SiciLlANO (1900),

Pacano (1900), Moissejeff (1927), Heymans (1928), Heymans
and Bouckaeut (1930). Through the investigations carried out
by Heymans and his co-workers it could be sho^vn that hypoxemia
does not directly affect the respiratory centre, hut acts almost
exclusively reflexly over the sinus and aorta regions. Of these
two reflexogenic zones, the first is by far the more important,
as is shown by tlie fact that only denervation of the sinus regions
inhibits the stimulating effect of hypoxemia upon the respiration
almost completely. Carbon dioxide, on the other hand, still has a
stimulating effect on the respiration after a severing of the nerve
connections from the aorta and sinus regions, so that we must
assume that carbon dioxide has an in the main direct effect upon
the respiratory centre.

As has already been mentioned, the localization and the great
physiological significance of the chemical receptor fields in the
carotid sinuses have engaged the interest of the majority of writers
in this sphere, while a comparatively small number, on the other
hand, have devoted a closer study to the reflexogenic, chemo-
sensitive aorta zone. Also the occurrence of chemical receptor
fields in the aorta has been questioned by some writers. Beyne,
Gauthelet and Halpern (1933) carried out experiments on dogs
anaesthetized with chloralose and enclosed in a chamber in which
the atmospheric pressure was then lowered. While the animals
could easily adapt themselves to tlie ventilation magnitude at a
sub-pressure corresponding to 8000 — 9300 meters, this was not
possible after bilateral severance of the nerve connections to the
carotid sinuses. Cyanosis then appeared rapidly at a pressure
corresponding to 3000 — 4000 meters, with cessation of the
respiration and death. The authors therefore denied that
anoxemia could exert any stimulus on the respiration other^v^se
than over the sinuses.

On careful denervation of the carotid sinuses in dogs, cats and
rabbits it sometimes happens that the respiration ceases and the
animal dies (W itt, Katz and Kohn 1934, Gemmill, Geiling and
Reeves 1934). Euler and Liljestrand (1936) as well as
Dautrebande and Wegria (1937) found in dogs and cats an-
aesthetized with chloralose that in the reflex effect on the

respiratory centre in connection with hypoxemia the earotid
sinuses play the main role, whereas the region that is innervated
by the depressor nerves is l^racticaIly insensitive to hypoxemia.

Other writers, again, have eonfirmed the existence of extra-
carotid chenioreceptors by showing that even after complete
sinus denervation hypoxemia still causes a slight increase of the
respiration, though tliis disappears when the vago-depressors are
severed. SELLADunAi and Wright (1932 a) studied the effect upon
the respiration at various oxygen concentrations on cats that had
been decerebrated or anaesthetized with chloralose, and found
that the respiration-stimulating effect of the hypoxemia was
entirely dependent upon the buffer nerves (.tltc depressor nerves,
the sinus nerves), hut that the latter are of far greater importance
than the afferent fibres in the depressor nerves. Schmidt (1932)
arrived at the same result in experiments on dogs and cats, hut not
on rabbits, in which he could not find any release of reflexes from
the aorta in connection with hypoxemia. Jonghloed (1936)
studied the regulation of the respiration in dogs in connection
with hypoxemia that had been provoked by lowering the
atmospheric pressure in a chamber. He found that after both
the depressor nerves and the sinus nerves had been severed there
was still a certain regulation of the respiration, though this
disappeared entirely if also the vagi were severed. This may be
explained by the fact that depressor fibres also run in the vagus
(Koch 1931). After severance of the sinus nerves and the vago-
depressors (as a rule already after severance of only the sinus
nerves) hypoxemia no longer stimulates to increased respiration,
but has rather a paralytic effect. These peripheral chemoreceplors
in the carotid sinuses and the aorta thus to some extent protect
the respiratory centre and thereby the entire body from the
consequences of hypoxemia. Also Gesell and Moyer (1937)
observed the existence of chemoreceplors in the aorta of dog in
experiments on hypoxemia. Lambert and Gellhorn (1938)
studied especially the effect of hypoxemia on the blood pressure
in narcotized dogs with pneumothorax and constant artificial
respiration. They gave gas mixtures poor in oxygen that in the intact
animals caused a rise in the blood pressure, but that in animals

in which the sinuses had been denervated and the vago-depressors
severed produced a lo^vering of tlie blood pressure. On a closer
analysis of how the effect was localized to tlie respective buffer
nerves the authors found that only the severance of the one pair
of buffer nerves did not cause a lowering of tlie blood pressure
in connection with hypoxemia. These experiments thus indicate
that the heightening effect of hypoxemia on the blood pressure
is dependent upon tlie chemoreceptors both in tlie carotid sinuses
and in the aorta region. CoMROE (1939) carried out a physiological
and anatomical investigation on the significance and the localization
of the chemoreceptors of the aorta in dog and cat.

The existence of other chemoreceptors tlian those in the carotid
sinuses and at the beginning of the aorta with any physiological
significance has not been demonstrated, despite the fact that
there exist so-called paraganglia with an appearance similar to
these in other places in the body.

n. Anatomical survey

The aortic and pulmonary paraganglia.

The occurence of limited cell groups of a paraganglionic nature in the
tract around the aorta and arteria pulmonalis has long been known, and has
been made the object of detailed studies by a number of morphologists. By
paraganglion is meant a cell formation which may be of either sympathetic
or parasympathetic nature. The former are developed from sympathetic, the
latter from parasympathetic neuroblasts. Also mixed paraganglia occur. The
existence of similar paraganglia in the carotid sinus region has also long been
known, but through Heiunc’s (1924) discovery of the sinus neiVes attention
was once more directed to this region, and a large number of researchers
(Druneh, 1925, De Castro, 1926, 1927, Riecele, 1928, Sunder-Peassmann,
1930, Boyd, 1937, and others) made it the object of investigations. Through
the experimental investigations carried out by Heymans and Heymans (1927)
and Heymans et al. (1930) data were obtained concerning the function of
these formations, that had previously only been the object of guessing and
free speculations, and also the physiologists had their attention dratvn to
problems in this field.

WiESEL (1906) described in connection with a morphological investigation
of the heart in children the occurrence of a limited, greyish red formation
embedded in the epicardial adipose tissue around the left coronary artery
trhere this runs medially to and behind the left auricle. It proved to consist
of typical chromaffin cells. Also Trinci (1907), in an investigation on
mammals and reptiles, observed the occurrence of chromaffin tissue around
the base of the heart. Busachi (1912) was able in two fully developed
embryos to distinguish two isolated groups of cells, an upper group localized
just below the aortic arch and a lower one, that already found by Wiesel
around the coronary artery. The upper group was characterized by deficient
chromaffinity, the lower, 'consisting of chromaffin cells, was not to be found
in adult persons. Rabl (1922) found in embryo of the guinea-pig similar
chromaffin formations laterally to the common carotid artery, between the
vessel and cervical sympathetic trunk. Especially on the right side they
occurred in large numbers, extending there down to the base of the heart.
He suggested the designation paraganglion caroticum inferius, in conformity
with the nomenclature introduced by Kohn (1900) (’’Nebenorgane des
periferischen Nervensystems”).

Watzka (1930) and Penitschka (1930, 1931) revived with their investiga-
tions interest in these cell groups of a paraganglionic nature. The latter
studied especially the paraganglion, which was situated below the arch of
the aorta in connective tissue between the aorta and the pulmonalis (Busachi).
To this formation, which was characterized by deficient chroinaffinity and
which he found regularly both in man and in mammals, the author gave the
name paraganglion aorticum sttpracardialc. It consisted of a cell-rich tissue
that was in part embedded in a rich network of vagal and sympathetic fibres,
where isolated ganglion cells also occured, and the cell groups lay in part
close beside the nerve branches, yet seemed always to be connected to the
ner%'es and through these to he bound up into an organic unit. He pointed
out the morphological agreement between this paraganglion and ’’paraganglion
caroticum”, although the former is smaller in size, and also mentioned that
these two formations are situated in regions from which the afferent regu-
lators of the blood pressure (the depressor nerve and Bering’s nerve) proceed.
They differ from the sympathetic paraganglia (paraganglion suprarenale, p.
aorticum abdominale) in that they do not consist of chromaffin cells and
do not produce adrenaline. The terms "paraganglionic” and "chromaffin”
have thus not the same meaning, and paraganglionic has a wider application,
for the paraganglia may consist of chromaffin (adrenaline-containing) cells
or of non-chromaffin cells (not adrenaline-containing) or again of a mixture
of these two kinds of cells. The sympathetic paraganglia disappears also in
older persons, with the exception of the suprarenal medulla (Zuckerkandl
1901, IWANOFF 1925), which is not, however, the case with paraganglion
aorticum supracardiale and ’’paraganglion caroticum”. The former are supplied
through sympathicus, but the latter are innervated in part through sympa-
thicus but in part also through the cerebral nerves. The paraganglion de-
scribed by Wiesel must therefore be considered to be of sympathetic nature.

Palhie (1934) carried out a careful histological investigation of the
paraganglia of the heart, and found that these, as Busachi had already pointed
out, could be divided into two groups, and he proposed for them the
designation paraganglion supracardiale superius (Busachi, Penitschka) and
inferius (Wiesel, Trinci). In the cat he did not find these two easily
distinguishable paraganglia, there was here a more diffuse distribution of
paraganglionic tissue, which extended along the aorta from the point of
departure of the left coronary artery and upwards to the egress of the large
vessels and to the point of attachment of ligam'entum Botalli, and along
pulmonalis from its commencement to its point of bifurcation. It was,
however, possible to observe a certain regularity in the localization in so far
as in definite places, namely, under the aortic arch behind ligamentum
Botalli, as well as just under the right pulmonary artery in the connective
tissue between the aorta and pulmonalis and also, finally, at the base of the
heart between the two main vessels, there were large groupings of cells. The
paraganglion situated under the arch of the aorta did not contain any

chromaffin cells. The other paraganglia, on the other hand, are nii.ved, hut
with mainly ehroniaffin tissue, and these arc found only in sonic full-grown
cats. A wealth of nerves run around these paraganglia. Very often paragang-
lionic cell-islands are enclosed in the nerves. In the aortic arch around
ligamentum Botalli and at the point of bifurcation of the pulmonary artery
the author found a remarkable wealth of vasn vasorum, which ran into the
media and which in the course of their passage through the adventitia some-
times received thin ner%’cs containing paraganglionic cells. The adventitia
of the aorta and pulmonalis are very rich in these nerves with paraganglionic
cells. Sometimes one also finds them in the outer layers of the media. He
assumed that they constituted the terminal branches of nervus depressor,
which c.xtcnds right into the media (KoESTEit and Tscheumak 1902). In new-
born cats he found that the upper paraganglion described by him sometimes
received its blood supply through a small vessel proceeding from pulmonalis.
A similar arrangement could sometimes be found in human entbryos.

Seto (1935), in an investigation carried out on adult human beings, also
found permanent paraganglionic tissue at the commencement of the aorta
as well as the paraganglia between the aorta and the pulmonary artery, which
he considered to receive a mainly parasympathetic innervation and only in a
lesser degree sympathetic. This innervation agrees with the investigations of
other writers on glomus caroticum (De Castro, 1926, RtECELE, 1928, Meijung,
1938). He considered that the afferent impulses from these formations must
be conducted in the depressor nerve, which, as is known, is characterized by
marrow-containing, centripetal fibres and has its region of distribution in
the tract for these paraganglia. He pointed out, further, how each paraganglion
lies close to a fairly strong main artery, or that such an artery runs to the
middle of it. The main artery divides in the paraganglia and sends branches
between the groups of cells to end, finally, in a number of capillaries that
surround the parenchymal cells. As was the case >vilh the arteries in the
paraganglia of the carotids (De Castro, 1927, Riegele, 1928), also the vessels
in these cardiac paraganglia showed a large number of sensitive nerve-
endings, which here doubtless derive from the depressor nerve.

Muratori (1934, 1935) described the occurence in mammals of other than
the paraganglia mentioned here above, which were situated, on the right side,
in the angle between the subclavian artery and the common carotid artery
and on the left side just above the aortic arch medially to the left subclavian
artery.

This discovery has been verified by Nonidez (1935), who observed in rabbits
and cats, in the places indicated by Muratori, the existence of this paired
organ, that consisted of two groups of round, non-chromaffin cells. These
organs were enclosed in a thin capsule of connective tissue. Exactly similar
cells were also found scattered around the endings of nervus depressor on the
corresponding side, and sometimes also embedded in the adventitia of the
aorta. The endings of the depressor nerve are found here in the adventitia.

Kipirea cor-
rcspondm}?
to those m
fi*r. 1

WlESEL

(1906)

TniNci

(1907)

Busachi

(1912)

Rabl

(1922)

Penitschka
(1930, ini)

Palme

(1934)

Localization

Around point of departure
for left coronary artery

Around base of heart

a) Just below aortic arch

b) Around left coronary
artery

Laterally to the common
carotid artery and dotvn to-
wards the base of the heart

Below aortic arch between
aorta and pulmonalis

a) Below aortic arch
between aorta and
pulmonalis

b) At beginning of left
coronary artery

Homo

(child)

Mammals

Reptiles

Homo

(full-time

embryos)

Homo

Mammals

Homo

(embryo)

Paraganglion

supracardiale

superius

Paraganglion

supracardiale

inferius

Muiiatori
(1934, 1935)

Nonidez

(1935)

Boyd

(1937)

Nonidez

(1937)

Nonidez

(1937)

a) Below aortic arch Homo
between aorta and
pulmonalis

b) At beginning of aorta

a) In angle between right Mammals
subclavian artery and
common carotid artery

b) Just above aortic arch
medially to left sub-
clavian artery

a) In angle between right Rabbits
subclavian artery and Cats
common carotid artery

b) Just above aortic arch
medially to left sub-
clavian artery

a) At brachiocephalic trunk Homo
just under point of Dogs
departure for right sub-
clavian artery

b) On ventral side of aortic
arch somewhat below the
point of departure for
left subclavian artery

Below aortic arch between Dogs
aorta and pulmonalis

Glomus

aorticum

Aortic body

Tab. 1. The table gives in summary form the paraganglia described in detail
in the text.

Fig. 1. Schematic drawing of the localization of so-called paraganglia around
the aorta and the pulmonary artery. For more detailed description see the
text and Tab. 1.

Imt also in the media of the right subclavian artery and the arch of the aorta,
thus corresponding to the site for the above-mentioned formations (Tschern-
JACHIWSKY, 1929). For this organ he introduced the term glomus aorticum.
He also points out that both the depressor nerves receive sympathetic fibres
mainly from ganglion cervicale inferius. That the depressor nerve contains
sympathetic fibres has long been observed (Schumacher, 1902, Tello, 1923,
and others). He found that the right glomus aorticum obtained its blood
supply through a small artery proceeding from the corresponding subclavian
artery or from the brachiocephalic trunk. Tlie left glomus aorticum was
supplied by an artery proceeding directly from the aortic arch or from the
left subclavian artery. These arteries contain in their walls a plentiful supply
of nerve-endings of baroceptor type.

In a later work by Nonidez (1936) he found in the kitten the small artery
already described by Palme which supplies paraganglion aorticum supra-

lordialc (Penitschka) and proceeds from somewhat varying parts of the
pulmonary trunk or from the right pulmonary artery, though as a rule from the
medial side of the trunk and supplies the paraganglion with venous blood.
In consideration of this blood supply the author suggested that the designation
glomus pidmonalis should he introduced instead. He pointed out that it
also probably received blood from the aorta or its branches.

Hollinshcad (1939) studied the nerve supply to glomi aortici in degenera-
tion experiments on cats, and found that the nerve fibres to glomi came
entirely from cells localized in ganglion nodosum. When ganglion cervicale
superius was removed by operation the author could not find any noticeable
change in the innervation.

Noniuez (1937) found that the localization of the paraganglia in dogs
differed from the localization in rabbits and cats. On the right side glomus
aorticum consisted of an extremely small group of epithelioid cells embedded
in the adventitia of the brachiocephalic trunk just under the point of
departure for the right subclavian artery. Thus the site for the paraganglionic
tissue in the human embryo described by Boyd (1937). On the ventral side of
the aortic arch somewhat belo%v the point of departure of the left subclavian
artery the left glomus aorticum is situated. It receives its blood supply
through a small artery proceeding from the ventral side of the aortic arch.
Branches from this artery also reach down to similar groups of epithelioid
cells situated on the ventral side of the place of bifurcation of the pulmonary
artery at the base of ductus arteriosus. In this artery there are nerve-endings
of baroceptor type. Such regions with baroceptors have also been
demonstrated in vena cava and in vena pulmonalis by Nonidez (1937). Here,
however, there are no paraganglia connected with them. Also in the space
between the aortic arch and pulmonalis the author described in the dog the
occurence of similar cell groupings, which probably correspond to
Penitschka’s paraganglion aorticum supracardiale in man. For this paragang-
lion he introduced the designation aortic body. It was supplied with blood
through an artery which proceeded from the dorsal side of the aortic arch.
A further group of cells of this kind lay between aorta ascendens and
pulmonalis, and corresponded to Palme’s paraganglion supracardiale
inferius. They obtained blood through a branch of the left coronary artery.
CoMROE (1939) was unable to find this group of cells in adult dogs. The
artery described in an earlier work of Nonidez (1936), which in the kitten
proceeded from truncus pulmonalis and supplied the paraganglia between
the aortic arch and pumonalis, was not to be found in the dog, and is thus
presumably not to be considered ns of any fundamental importance, and
the author accordingly abandons his previous opinion. In a work by
Goormachtieh and Pannier (1936) it was pointed out that this artery is

only to be found in the embryo of the cat and in the kitten, but is obliterated
in the full-grown animals, in whom this paraganglion is thus supplied with
blood through the coronary vessels or aorta ascendens.

In connection with serial sections Addison and Comhoe (1938) found that
in dogs the paraganglionic grouping of cells described by Pcnitschka and
situated in the adventitia of the aorta obtained its blood supply through a
small vessel proceeding from the aorta on a level with the brachiocephalic
orifice.

The approximate positions of the paraganglia are marked in a schematic
drawing (Fig. 1).

AUTHOR’S OWN INVESTIGATIONS

III. Aortic body

1. Technique and Procedure.

Cals were used in most of the experiments but in some cases
rabbits irere employed. The cats were anaesthetized Avitli
0 . 03 — 0 .O 7 g of chloralose per kilogram of body-weight. The
animals were first anaesthetized with ether, after which chloralose
was injected intravenously in a 1 % solution. The rabbits were
anaesthetized •with urethane, 1.4 g per kilogram of body-tveight
being injected intravenously in a 20 % solution.

A tracheal cannula was introduced in the usual way. The blood
pressure was recorded from the femoral artery by means of an
Hg-manometer. Intravenous injection -was given through tlie
femoral vein.

In these experiments the respiration was recorded with an
ordinary Marey pneumograph connected mth a Marey tambour.

For local, intra-arterial or intracardial injection a fine ureteral
catheter was employed, which was moistened with glycerin to
ensure an easier passage through the vessels. The catlieter was
generally introduced into the circulatory system through the right
common carotid artery in tlie neck. In some cases it was intro-
duced via the right or the left subclavian artery. By pushing
the catheter do^sm carefully it was possible to get the orifice of
the catheter to lie in the spot desired. After every experiment
the exact position was determined on autopsy. In order to get
the catheter to glide through aorta ascendens it proved best.

Giving to the anatomical conditions, to twist the animal, which
was lying on its hack, hard to the left. The passage past the
aortic valves was recognized by a slight resistance, -which was
easily overcome. It was not possible to observe any considerable
tearing of the valves. Different substances could then be injected
through the catlieter (lobeline, piperidine etc.).

In those cases in tvhich a recording of tlie action potentials
from the depressor nerve was carried out the dissection procedure
was as folJotvs. In cats the skin and musculature were dissected
away over the clavicle and the three or four uppermost ribs
(generally on the right side), which were severed. If the inter-
costal arteries and the internal mammary artery were ligated this
operation could be carried out without any bleeding worth
mentioning. After this tlie depressor nerve was sought at its point
of entry into vagus, tvhich was in the tract under the clavicle,
and dissected out in a caudal direction. This was done under a
magnifying glass, care being taken to see that the nerve was not
stretched or pinched. The nerve was severed as far cranially as
possible, and in order to increase the signal-to-noise ratio the
nerve-sheath was carefully drawn off. To prevent the nerve from
drying, it -was kept moist -Huth Ringer’s solution at body-tem-
perature, which was dropped onto it with a fine pipette. During
the registration the animal was, moreover, placed in a moist
chamber.

The recording of the action potentials was made by means
of a capacity-resistance coupled amplifier and a cathode ray
oscillograph. This amplifying set allowed an amplification of up
to 10' times, and thus enabled me to record the action poptentials
from very thin nerve fibres, provided tlie nerve preparation
between the electrodes was of a small diameter, thus offering a
high signal-to-noise ratio (see Zotterman 1936). The potentials
were led off by means of Ag-AgCl-electrodes of the conventional
type.

In tlie experiments on cats where the action potentials were
recorded one was compelled, since the thorax was open, to give
the animal artificial respiration by means of a Starling pump.

Fig. 2. Localization of the chemoreceptors of the aorta in dog. Right top:
microphotography of the aortic body with its aortic branch. Below is seen
the effect, in connection with electric stimulation, on the respiration and
blood pressure of the afferent branch to the aortic body (J. H. Comroe: Am.
J. Physiol. 1939, 127 , 176).

By connecting the pump with bags containing various gas-
mixtures it was possible to study the effect of these on the action
potentials that were elicited by chemical stimuli.

2. Localization.

Since in their fundamental work J. F. Hetmans and C.
Hetmans (1927) showed that impulses stimulating the respiration
could be elicited through chemical stimulus on perfusion of the
heart and the nearest part of the aorta of dog wth the head,

Fig. 3. Schematic drawing of the localization of certain so-called paraganglia
(in part after Nonidez; Amer. J. Anat., 1935, 57, 259). The figures 2 and 3
refer to the paraganglia described by MuBATOni and Nonidez; and 1 marks
the aortic body. I, 11, III and IV mark various catheter positions. For further
explanation see text.

apart from the vago-depressora, entirely isolated from the body,
no new attempt at a closer study of the exact situation of this
chemo-sensitive zone was carried out until CoMROE (1939) tried
to locate the position more exactly in dogs and cats, taking a
physiological line of approach. By means of local application
in the aorta of dogs Avith denervated carotid sinuses of certain
stimulants (cyanide, lobeline) through injection via a fine
catheter, he Avas able to get an idea of the spot Avhere the strongest
reflex excitation Avas released. With this procedure, combined
Avith local denervation, to Avhich I shall revert later, he Avas
ahvays able to localize the chemo-sensitive region to aorta
ascendens or to the beginning of the aortic arch (Fig. 2). This
chemo-sensitive region of the aorta Avill be referred to in the

Fig. 4, Cal 2.2 kg. After bilateral sinus denervation. The upper curve blood
pressure, lower curve respiration (pneumograph). At the arrows injection
through a catheter of 0.3 mg of lobeline hydrochloride. In A the tip of
the catheter is placed just inside the aortic valves. In B the catheter tip
lies outside the aortic valves in aorta asccndens.

sequel as the aortic body, under which name it is most commonly
known in the literature.

If the opening of the catheter was placed just inside the aortic
valves (Fig. 3, catheter position I) of an animal with denervated
carotid sinuses one obtained a prompt increase of respiration on
injection of 0.3 mg of lobeline hydrochloride (Fig. 4A). This is
in full agreement with Comroe’s experiments on cat. With a view
to tlie paraganglia described by Muratori and Nonidez the
catheter %vas introduced through tlie right or left subclavian
artery so that the opening came to lie in the spot from which
the vessels proceeded which supplied these formations with blood
(Fig. 3, catlieter positions II and III). But in these positions no

increase at all in tlie respiration was obtained on injection of
lobeline, even in doses far exceeding that given above. Bearing
in mind the possibility that owing to the blood stream in these
catheter positions the lobeline did not reach these paraganglia,
lobeline was injected in aorta ascendens via the catheter (Fig. 3,
catheter position IV) but still without any increase in the respira-
tion (Fig. 4B). One arrived at the same result also tvhen the
action potentials were recorded from the depressor nerve. In
catheter position I one obtained a strong increase in the chemical
impulses on injection of lobeline, but this increase was entirely
absent in catheter positions II, III and IV. It wotild thus appear
tliat from these paraganglia one cannot elicit any reflex effect
on the respiration with a chemical stimulus. On the other hand,
it is easy to produce such an effect from the paraganglia described
by Penitschka (paraganglion aorticum supracardiale) (Table 1
and Fig. 1). Further proof of this opinion will be given below.

3. Nerve supply.

Since Heymaks and Heymans showed that the afferent impulses
which arose on chemical stimulus of tlic aortic region in the
dog passed in the vago-depressors, this has been verified by a
large number of investigators, and it is notv generally accepted
as a fact. CoMfiOE (1939) found, as was expected, that in dogs
with denervated sinuses no stimulating effect could be obtained
on the respiration and blood pressure on stimulation with
hypoxemia, lobeline or cyanide if the vago-depressors were
severed. On the other hand, an effect could be produced wdth
these stimuli if the vagi were severed below the point of entry
of the depressor nerves. This shows, as was already knoA™, that
the afferent path for the impulses from glomus aorticum is the
vagi via the depressors.

On dissection of the depressor nerves in cat it proves that these
join the vagus trunk immediately beloAv nervi recurrentes or
together Avith these. But from the depressor nerA'es there proceed
in the majority of animals on both sides one or several branches

wliich leave the main trunk after approximately half its course
from the terminal branching to the point of entry into the vagus.
These branches, which are extremely tliin, turn off more sharply
laterally and join tlie vagus 1 to 2 cm below the point of entry
of the depressor nerve proper. Such thin nerve fibres between
the depressor nerve and vagus have been described by Pehman
(1920), but were then localized somewhat higher up. If one
records the action potentials from these ’’accessory” branches
from the depressor nerve one gets the same picture as from the
main trunk of the nerve, thus the ordinary fast pressor impulses
at each systole as well as the slower chemical potentials elicited
by e.g. hypoxemia or different drugs (lobeline, piperdine etc.).
These ’’accessory” branches from the depressor nerves may pos-
sibly provide the explanation of the fact that in cases of
hypoxemia the effect on the respiration after severing the carotid
and depressor nerves to some extent remains, and is not entirely
lost vmtil also the vagi are severed (Jongbloed 1936, Lambert and
Gellhorn 1938). Presumably only the main trunk of the de-
pressoi: nerves, which is from the technical point of view easier
to get at, has been severed, while impulses from the chemo-
sensitive zone of the aorta have been able to pass in the ’’acces-
sory” branches, which join vagus further down, and have then
passed up and thus produced an effect on the respiration.

In accordance with the theory advanced by De Castro (1927),
Heymans and Boucicaert (1933) were able to show that the
receptor organs for the chemical stimulus in the sinus region
are not identical with or localized to the same spot as the baro-
ceptors. CoMROE (1939) , in one experiment on dog with denervated
carotid sinuses, was able to destroy mechanically the nerves
running in the vicinity of the aortic body, and thus prevent the
release of the chemical reflex, while the pressor reflex remained
unchanged. This is in full agreement with the experiments carried
out by Bouge and Stella (1934, 1935), and by Zotterman (1935).
In experiments on the carotid sinus, with recording of the action
potentials from Hering’s nerve by severing or squashing the nerve
fibres from the chemo-sensitive or pressor-sensitive regions, these
authors %vere able to obtain either chemical or pressor impulses.

Taking as the point of departure these earlier investigations,
it lias been possible, in connection with the recording of tlie
action potentials in the depressor nerve, to make a closer study
of the result of a local destruction of the terminal branchings of
this nerve in the region around the concave side of the aortic
arch. By squeezing tsdth a pair of tweezers in the region between
aorta ascendens and tbe pulmonary artery, thus tlie site of the
aortic body, it was fairly easy to destroy the fibres leading tlie
afferent impulses from the chemo-sensitive region. By subsequent
checking of the effect of loheline or hypoxemia on the chemical
action potentials in the depressor nerves it was possible to
convince oneself that these were now no longer obtained, but that
the pressor impulses remained apparently unchanged. In one
case an involuntary exclusion of tbe chemical impulses from a
previously good subject was obtained, when in connection with
an intravenous injection of lobeline tbe cat gave a jerk, so that
the lifted depressor nerve, which tvas tied with a silk thread with
a fixed point of attachment, and from which the action potentials
were recorded, was over-stretched. After this, no further chemical
impulses were obtained, while on the other band, pressor impulses
came as before. But I never succeeded entirely in only destroying
those nerve fibres which conducted the presi-or impulses, without
at the same time more or less destroying the nerve fibres from
the aortic body. Since the pressor impulses, as compared with
the chemical impulses, are released from such a relatively large
region, there are great technical difficulties connected with the
destruction of tliese without the chemical fibres becoming at
the same time involved in the process. These experiments,
however, support the assumption that tbe chemoreceptors in the
aorta as well as in the carotid sinuses are anatomically distinguish-
able from the baroceptors.

In view of the fact that in rabbit tbe main trunks of the
depressor nerves run isolated from the vtigi, this animal should
provide better conditions for investigations of the chemo-sensitive
aorta zone. One 'would in this case be able to record the action
potentials from the nerve where it runs in the neck, without
being compelled in tbe majority of cases, as one is witli the cat.

mnaw/wAAA/wvwvmvwwvvvvvwvwvv^

Fig. 5. Babbit 2.4 kg. Action potentials from tlie right depressor nerve.
A. Control. B. After intravenous injection of 1 nig lobeline hydrochloride.
Time: 1/50 see.

to make a resection of the uppermost ribs on one side and on
this account being forced to administer artificial respiration to
tlie animal throughout the experiment. In exceptional cases,
however, one finds tliat also in cat the depressor nerve runs
isolated from vagus for a longer or shorter stretch in the neck
instead of, as in most cases joining the vagus already in the thoracic
cavity. It no^v proved, however, that on chemical stimulus of the
presumed chemo-sensitive aortic zone in rabbit and with simul-
taneous recording of the action potentials from the depressor
nerve no change at all of the electro-neurogram could be obtained
to indicate the existence of chemical impulses (Fig. 5 ). In order
to produce a chemical stimulation recourse was had as usual to
lobeline, cyanide or piperidine (intravenously or locally by
intracardial injection) and to hypoxemia varying from 12 % to
6.0 % Oo in N2. If these animals were allowed to breathe a gas
mixture poor in oxygen (7.3 % O2 in N2) during the recording
of the respiration with a Marey pneumograph, one obtained the

Fig. 6. The same rabbit as in Fig. S. Registration of blood pressure. Tbe
first arrow indicates tbe inonicnt when tbe clectrn-ncurogram A in Fig. 5
is taken. The second arrow indicates the point of time for tbe injection of
1 mg lobclinc hydrocldoridc intravenously. Tbe third arrow marks tbe point
of time wlicn tbe electro-ncurogram B is taken.

ordinary increase in tlic respiration. If after this one severed the
two depressor nerves and repeated the experiment one obtained no
measurable change in the increase in the respiration. On intrave-
nous injection of 1 mg of lobeline hydrochloride the case tvas the
same, namely, that the severing of the depressor nerves did not
reduce the stimulating effect on the respiration. In rabbit, the
intravenous injection of lobeline in doses of O.s — 1 mg always
caused a reduction of the blood pressure (Fig. 6) instead of, as
mtli the cat, causing a rise in the blood pressure.

Not all the fibres of the depressor nerves in the rabbit run in
complete isolation from vagus, they also sometimes occur to some
extent together with the latter. This is also observed if one splits
up the vagus in thin fascicles and records the action potentials
from one of these. In this case one can sometimes obtain a fascicle
containing depressor fibres, which may be recognized as conduct-
ing the fast impulses that are elicited by each pulse wave. It is
therefore possible that the afferent, chemical impulses might
travel in the fibres that were enclosed in the vagus trunk or in
vugus and not in the isolated depressor nerve. Any attempt to lead

i i n ■ • • ' ■ ... - < • ■

\aaaaaaaA/www\aaaaA/^

1 !

kAA/\7VV^^

Fig. 7. Ralii)it 2.0 kg. Recording of the action potentials from a thin fascicle
of the right vagus. A. Control. B. After spraying of trachea and right
bronchial system with 2 % pantocaine solution. C. After intravenous injection
of 0.75 ing of lobelinc hydrochloride. Time: 1/50 sec.

off tltese specific impulses from the vagus after chemical stimula-
tion of the aortic body presents, liowever, as already stated,
great difficulties on account of the very strong volley of impulses
that is released with each breatli, Gehnandt and Zotteeman
(1945) met with these difficulties in connection with their
attempts to find afferent vagus fibres that responded to increased
pressure in the central veins (the Bainbridge reflex). By means
of various procedures they attempted to eliminate the afferent
fibres of the lungs. The best result was obtained by introducing

blood was used for estimation of histamine in plasma. The skin
samples weighed 25 — 15 mg. The results, summarized in table
4, agree with those of table 3.

TABLE 4.

Rat ar

Date

19.4

3.5

30.5

17.5

19.4

2.5

24.0

Plasma Hi, y/1

140

14.5

150

250

225

150

165

150

Skin Hi, y/I

24.8

27.9

22.3

30.0

32.0

35.5

19.5

16.5

20.5

Rat nr

Date

18.4

17.5

20.4

2.5

19.4

17.5

17.4

3.5

Plasma Hi, tl\

190

220

315

300

145

165

Skin Hi, v/i

10.9

15.0

33.7

37.5

32.8

28.2

31.9

27.0

c. , Experiments on dogs. The histamine content of plasma
was determined in two dogs. In an ear vein of a nonanaesthe-
tized dog an incision was made from which about 20 ml of
blood w\as collected in a centrifuging tube, containing heparine.
The sample was treated as in the previous experiments. Table 5
demonstrates the constancy of the plasma values.

TABLE 5.

Dog nr

Date

20.2

2,3

15.3

7.4

27.2

8.3

18.5

Plasma Hi, y/1

d. Experiments on rabbits. 10 — 15 ml of blood was collected
from an ear vein of the nonanaesthetized animal. From table 6
it is obvious that the histamine concentrations of the plasma
samples show very great variations while the histamine content
of blood remains fairly constant.

In this respect the rabbit obviously differs from the other

lobeline the same view has been arrived at in these investigations
on recording the respiration and the action potentials from the
depressor nerve. A catlieter was introduced in the usual way
through the common carotid artery into the left ventricle of cats
with denervated carotids. If the opening of the catheter was
placed inside the valves of tlie aorta and lobeline was injected
one obtained a prompt effect on the respiration and blood pres-
sure. If the catheter was placed in such a way that the opening
lay distally to the valves, this effect on the respiration did not
appear (Fig. 4).

On the injection of cyanide or lobeline in the right ventricle
CoiMROE shelved that there was always a longer interval between
the injection and the appearance of an effect on respiration and
blood pressure than after the same injection in the left ventricle.
This would support the assumption that the chemoreceptors in
the cat, as also in the dog, do not obtain blood from the pul-
monary artery.

these two different types of action potentials on recording from
the sinus nerve. The big ones were due to pressure variations in
the sinus, the small ones were caused by the changes in chemical
composition of the blood. The latter kind of impulses, which
were of the continuous type, proved to have a considerably lower
frequency (about 20 — 40 per second) than those which were
released by the endosinusal pressure variation, which had a
frequency of about 100 per second. Tbe small impulses remained
even if the carotid artery was clamped below tlie sinus region,
so that the impulses produced by the pulse wave were eliminated.
The assumption of these authors, that tliese small action potentials
are caused by the chemical stimulus of the blood, was confirmed
by later investigations carried out by BouGE and Stella (1934,
1935) and Zotterman (1935). Taking as their point of departure
tlie already mentioned observations made by Heymans,
Bouckaert and Dautrebande, these authors succeeded in me-
chanically excluding the baroceptors, and were tbus able to get
a clearer picture, especially of the chemical action potentials.
Zotterman showed that the amplitude of these small potentials
were only 10 — 20 % of the largest spike potentials which were
elicited by stretch and that there was the same relation between
the amplitude of the taste potentials and the largest toucli
potentials recorded from the glossopharyngeal nerve running
to the tongue. Bouge and Stella found that on a stronger
stimulation of the chemoreceptors the chemical potentials were
considerably greater. This is presumably due to a summation
effect. These authors and Zotterman were able to show that the
number of chemical impulses increased considerably in connec-
tion with hypoxemia or asphyxia.

The assumption of Hey'mans and his co-workers that the
occurence of a chemical sensibility in the sinus and aorta regions
was bound to specific receptors distinguished from baroceptors
received strong support in De Castro’s (1926, 1927) morpho-
logical investigations on the receptors in the sinus region. This
view has since been confirmed by investigations carried out by
SCHMIUT (1932), Heymans and Bouckaert (1933), Gollwitzer-
Meier (1934), Bronk and Stella (1934) and others. The latter

"ir:-

S^^M^^ ^ ^geB ^nitaUO^ .g; (19391’ ° ,T\cd

, ten’ll '“"’"lioa »''“ ° J ZO«“"''’^ *c ca>’ “loUe

acU J i^vesUg an^ ^„ixW j.ate«i

a „u\aUO»' ''' ' tvpoxeT«^»’ By

-bVi “^“crS'a-' '■\"'“jr» r-

oxygen. gen ^ accotdt«6 {tom

aa .oon as ’n iens'° ^sct and

aetetmVn\^6^\^icit "^'^^anon o^ °^,yp.

tliey .aBy ^ n „att ^!t

*'“'°e™S» ’*»'.*%raa, a« '“t’r»*<>« '’t'^e””''

S'»- °“' “" ot 6i». a *' " o( a«« ”"1

*.l*e ^,„y B ««'“'” a« %0TtE»'

ii93«’ bio«i«6

elimm”'"' ’ ^ glwwel jeal 6*""’'

lorrtw'^" r .^.seaaya.e""
aoWl-"''-"'*’

• n Y>o(®”aa® ’.(.g ^ro5«ce rj^icsc

On,BCB.«n, t-1 .„,11 '’"9f;„gblereJ

a-* "Tir*=a.w

laltei mnst, i

Wm^WMMW\IW\f\IW\IW

Fip. 8. Cal 3.0 kp. Artificial respiration, constant over-ventilation witli air.
Recording of the action potentials from the right depressor nerve, A. Control,
BP 110 mm. Hg. B. 6.8 % oxygen in nitrogen, BP 120 mni. Hg. Time:
1/50 sec.

to be cansed by tbe chemical elimiibis of the blood, or in other
^vords to be elicited by hypoxemia, carbon dioxide or both
factors. By increasing the artificial ventilation, so that the carbon
dioxide tension was kept under the threshold value for stimidation
and the oxygen tension of the blood was increased, a clear
reduction of the small, chemical impulses was obtained.

After this, and with the same ventilation, the effect of gas
mixtures poor in oxygen, varying between 5.c to 12 % oxygen
in nitrogen, was tested. Fig. 8 shows the effect on the chemical
action potentials of ventilating tbe lungs with 6.s % Oo in N 2 -
One sees liere the typical increase of the action potentials produc-
ed by the chemical stimulus, which is in full argeement with
similar experiments on Hering’s nerve.

Fig. 9 shows the effect on the chemical action potentials from
the depressor nerve on ventilation with pure oxygen. This caused
an almost complete disappearance of all the small impulses.

yVWi/VVWVVt^^

‘ j II I . i^rli \ \ "'

Fig. 9. Cat 3.0 kp. Artificial respiration air. Recording of tiie action potentials
from the right dcpres.“or ncr%’c. A. Control, Bl’ 145 inni. Hp. B. 100 %
oxygen, BP 140 mm. Hp. Time: 1/50 see.

While the effect on respiration and blood pressure produced
by the hypoxemia goes almost entirely via the chemoreceptors,
tlie case as regards carbon dioxide is complicated by the fact that
a part of the effect is direct on the centre. Heymans, BouciCAEnr
and DAUTREnANDE (1930) were able in perfusion experiments on
dog to show from the sinus region, which was isolated in respect
of the circulation, that a rise in the carbon dioxide tension caused
a reflex increase in the respiration over the chemoreceptors. The
carbon dioxide still had, however, a stimulating effect on the
respiration after severance of the buffer nerves, so a direct effect
upon the respiratory centre had also to be assumed. Special
experiments, however, led them to adopt the view that the
sensitiveness to carbon dioxide of the rcflexogenic zones was
greater than that of the respiratory centre. As a result of further
study, Heymans and Bouckaert (1939) found that the effect
over the chemoreceptors is quicker in starting than that over the

centre, ^^•hose reaction is slower but longer lasting. Gollwitzeu-
Meier and Lerche (1940) observed also that carbon dioxide
was active over the chemoreceptors but they tvere on the other
liand of the opinion that the threshold value of the latter was
higher than that of the respiratory centre. The chemoreceptors
therefore had only an auxiliary function, which set in only when
liigher concentrations of carbon dioxide had been reached, but
within physiological limits.

Although it is difficult to draAv a comparison here, as an isolated
elimination of either organ can only be realized as far as the sinus
and the aortic bodies are concerned, it nevertheless emerges beyond
doubt from experiments by Euler and Ltljestrand (1940) that
the respiration was reduced after denervation of the reflexogenic
zones in the sinuses also after it had been possible to exclude
hypoxemia. The view advanced by Owen and Gesell 1931,
Schmidt 1932, Gemmill and Reeves 1933, Wright 1934, 1937,
Stella 1935, Gesell and Moyer 1937 and Smyth 1937, that
carbon dioxide has no effect over the chemoreceptors, is probably
in some way connected with tinfavourable experimental condi-
tions.

The real question here is whether under physiological condi-
tions carbon dioxide exercises on the respiratory centre any
stimulating effect ^vhatever that is reflexly elicited from the sinus
and aorta regions. Schmidt and Comroe (1940) have strongly
denied any such effect under physiological conditions.

Saaman and Stella (1935), recording the impulses from
Bering’s nerve, found that the threshold value for stimulation
of the chemoreceptors for carbon dioxide lay at a carbon dioxide
tension of about 33 — 35 mm Hg in the arterial blood. This agrees
^^’ell with the values obtained in the experiments by Euler,
Liljestrand and Zotterman (1939). They found also that such
carbon dioxide concentrations in the gas mixture supplied on
artificial ventilation which led to alveolar carbon dioxide values
entirely within the normal limits brought about a marked increase
in the impulse frequency. On increase of the carbon dioxide
content the impulse frequency rose in an approximatively linear
relation. They therefore considered that carbon dioxide also had

Fig. 10. Cal 3.0 kg. Artificial respiration, constant over-ventilation with air.
Recording of the action potentials from the right depressor nerve. A Control,
BP no mm. Hg. B. 6.3 % carhon dioxide in oxygen, BP 115 mm. Hg. Time;
1/50 sec.

a stimulating effect on the centre that was reflexogenically elicited
from the carotid sinuses.

It is probably now beyond doubt that under normal conditions
carbon dioxide exercises a reflex stimulating effect over the
chemoreceptors in the sinus and aorta regions, but that in the
absence of tbis the central effect on the respiratory centre is
still in a position to regulate the ventilation, although at a
liigber threshold level, as is shoyvn, moreover, by the fact that
apnoea may sometimes appear on denervation.

2. Hypercapnia.

In order to be able to study the effect of carbon dioxide on the
aortic body, the animal was kept in a slight over-ventilation with
an oxygen mixture in which carbon dioxide entered in different

concentrations. This ^cas to ensure that tlie animal should he
fully saturated with oxygen, so that all action potentials that were
caused by liypoxemia might be eliminated. Fig. 10 shows the
effect of 6..-! % CO.i in Oo. One sees the increase of the chemical
impulses, Avhich must be considered to be produced by the
increased carbon dioxide tension in the blood.

This is in full agreement with earlier experiments on the
afferent impulses in Bering’s nerve under similar conditions.

nn the chem«-

f various drugs on «

The effect ol tody

receptors m

i^obclinc.

,ue .n»uWta. eHec. ot ^

Jen k»o«- »;“ t « toealtoi to *0 t«P> i.eesng.«->'''

malted been e ^je e

it the caw"* ^ to do^ »' " aeneiva"”" ■> bo*

injeclion o btainedi '"“ togo-deriM^ aU*o»

„.p>t.u«n «» .eve«"«^“\ * n..TW■‘'‘^ “oitation

'“““n’”? J «• ‘’■‘"k* ot lobeline on ^„rt

aides *e “f' o,e .tintnlaling t"" o,id w"". bottont

'»““T It conditioned ovJ *» ,i„et to^ ^ „(

ismairvly 3^ too, a sUtnui

of the CourlU ve

Fig. 11. Cal 3.2 kg. Artificial respiration air. Recording of the action potent-
ials from the right depressor nerve. A. Control, BP 165 mm. Hg. B. After
intravenous injection of 1 mg lobeline hydrochloride, BP 190 mm. Hg.
Time: 1/50 sec.

the respiration. It ^s'ould thus appear that under certain circumst-
ances also the respiratory centre can be stimulated by lobeline.
Gollwitzer-Meier (1934) was also able to constate that lobeline
takes effect of the sinus region, but scarcely, on the other hand,
on the respiratory centre. Euler, Liljestrand and Zotterman
(1939), in connection with the registration of impulses from
Hering’s nerve, tsere able to show that lobeline causes a much
stronger increase of the chemical impulses than hypoxemia is
ever able to bring about.

In my experiments on the stimulating effect of lobeline on the
chemoreceptors in the aortic body the lobeline has been given in
part intravenously and in part intracardially (catlieter position I,
Fig. 3, p. 23). Fig. 11 shows the effect of 1 mg of lobeline hydro-
chloride intravenously upon the inflow of impulses in the right
depressor nerve. One sees how the lobeline causes a very marked

Fig. 12. Cat 3.2 kg. Artificial respiration air. Recording of the action potent-
ials from the right depressor nerve. A. Control, BP 160 nini. Hg. B. After
intravenous injection of 1 ing piperidine hydrochlorid, BP 190 nun. Hg.
Time: 1/50 sec.

increase of the chemical impulses, which gradually disappeared
in the course of some minutes. Also the fast and larger spike
potentials increased, but this was due to the rise in the arterial
blood pressure.

2. Piperidine.

Eulek (1945) showed that piperidine is normally excreted with
the urine, and that this excretion increases considerably on
muscular exertion. Its typical nicotine-like pressor effect made
it probable that it could stimulate the chemoceptive fibres in
the carotid bodies. Gernandt and Zotterman (1945) were able
to show that piperidine has an effect identical with that of
lobeline on the chemoreceptors in the carotid sinus.

Fig. 12 shows the effect of intravenous injection of 1 mg

Fig. 13. Cal 3.8 kg. After bilateral sinus denervation. Upper curve blood
pressure, lower curve respiration recorded with a pneumograph. At first arrow
injection via a catheter with the tip inside the aortic valves of 0.3 mg loheline
hydrochloride; at second arrow injection of 0.3 mg piperidine hydrochloride.

piperidine hydrochloride on the chemoreceptors in the aortic
body by recording the impulses from the depressor nerve. A
comparison witli Fig. 11 shows this effect of piperidine to he
identical witli that released by lobeline.

Fig. 13 shows the effect, on local, intracardial injection, of
0.3 mg loheline hydrochloride and 0.3 mg piperidine hydro-
chloride on the respiration and blood pressure of an animal
with severed sinus nerves.

3. Cyanide.

GepI’ERT (1889) pointed out tlint the toxic effect of the cyanides
was due to tJie fact that they caused an internal asphyxia. Conse-
quently, owing to a hlochnde of certain oxidation cnxymes, the
cyanides have the .same effect on the respiration as lias hypoxemia,
aiid variows researchers have also been able to show this.
Hetmaxs, Bouckaert and DAUTnEHANnE (1931) showed that the
stimulating effect of cyanide on tlm respiration was caused mainly
by a reflex effect from the carotid sinus, and that the sinus was
much more sensitive to this stimulus than xvas the respiratory
centre, which was practically insensitive. The same thing also
applies to the effect on the hlood pressure. OwEX and Gesele
(1931), however, found that cyanides could produce a certain
increase in the respiration even after complete bilateral sinus
denervation. Heymans, Bouckaert and Kegmers (1933) also
arrived at the same result. Wright (1933) observed that after
cyanides there appeared two pha.ees in the increase in respiration.
The first was rcflcx-conditioned and was presumably caused by
the hypoxemia occuring in the chemoreccplors; the second stimul-
ation of the respiration, Svhich lasted longer, avas not changed by
denervation of the sinuses. The latter xvas considered by Eueer
arid Liuestraxd (1937) to be due to the formation of acid
metabolites, which arose in consequence of the reduced oxidation
m the whole body and which caused a stimulation of the respi-
ratory centre and also of the chcmorcceptors if these were intact,

T'l * • • ^

us View is also supported by an investigation by Gesell,
Kreucer, Gorhaju and Rosenthal (1930), who found that intra-
'enoug injection of cyanide w’as follow’ed by a rapid increase of
the lactic acid in the blood as well as by a rise in the hydrogen-ion
concentration of the arterial blood.

On intravenous injection of Ringers solution containing O.o2%
0 potassium cyanide, with simultaneous registration of the action
potentials in the depressor nerve one obtains a clear increase of
' '0 chemical impulses. The electro-neurogram in Fig. 14B was
ta en minutes after the commencement of the infusion. The
ommal had then received 3.2 ml of potassium cymnide solution.

Fig. 14. Cat 2.9 kg. Artificial respiration air. The action potentials from the
right depressor nerve. A. Control. B. During infusion of a weak cyanide
solution. Time: 1/50 sec.

This is in agreement with Saaman and Stella (1935), who
recorded action potentials from Hering’s nerve and found an
increase of the chemical impulses after cyanide. Euler, Lilje-
STRAND and ZoTTERMAN (1939) found that if in connection with
the continuous supply of potassium cyanide solution, with simul-
taneous recording of the impulses from the sinus nerve, they
administered artificial respiration with pure oxygen instead of
air, the chemical impulses caused by the cyanide were reduced
very considerably or disappeared altogether. This is in full
agreement with the close resemblance between the effect of
cyanide and that of hypoxemia.

4. Acetylcholine.

Also this substance has a stimulating effect on the respiration
over the chemoreceptors in the sinuses (Heymans and Haadovsky
1935, Heymans,Bouckaert, Faber and Hsu 1936, Schweitzer and

/VVWWVVWWW^

V ^ ' 'I

jiWVVWW^^

Fig. 15. Cat 2.4 kg. Artificial respiration air. Action potentials from the right
depressor nerve. A. Control, BP 155 niin. Hg. B. .After injection via a
catheter with the tip inside the aortic valves of 10 pg acetylcholine, BP
120 mm. Hg.

Wright 1938, Winder 1938, Euler 1938, Comroe anti Schmidt
1938). Euler, Liljestrand and Zotterman (1941), in connection
with the intrasinusal supply of acetylcholine, showed how the
chemical impulses increased on recording the action potentials
from Hering’s nerve.

Fig. 15 shows the effect of 10 pg acetylcholine on local, intra-
cardial injection. One sees an undoubted increase of the typical
cliemical impulses.

VI. Selective elimination of the cliemorecep-
tors in the carotid and aortic bodies

In some previous experiments Zotterman and I observed that
strong acids applied intrasinusally could completely extinguish
the chemical impulses as was confirmed with recordings from
Hering’s nerve, while the pressor impulses remained apparently
unchanged. Euler, Liljestrand and Zotterman (1939), recording
the impulses from the sinus nerve of the cat, showed that after
intravenous injection of 0.5 to 2 ml of an 0.5 N ammonia solution
all the small impulses disappeared, while the big pressor impulses
still appeared as before. If, immediately after this, the animals
were allowed to breathe oxygen-poor or carbon dioxide-rich gas
mixtures, no increase of the chemical impulses whatever was
obtained. Also the ordinary increase in the flow of impulses due
to stimulation of the chemoreceptors on infusion of a weak
cyanide solution failed to appear after the injection of ammoniac.
This is in full agreement with the view that cyanides give rise
to hypoxemia. This failure of the stimulating effect was, however,
of extremely short duration. After the lapse of a minute or so the
chemical impulses returned as before.

Lobeline, nicotine (Heymans and co-workers), acetylcholine
(Heysians and Handovsky' 1935) and potassium ions (Euler
1938) have, as has been mentioned, a strongly stimulating effect
on the respiration and the blood pressure over the chemo-
receptors. They constitute a pharmacodynamically homogeneous
group, in so far as all are typical stimulating substances on the
transferance of stimulus from preganglionic to postganglionic
fibres in sympathetic ganglia (sjTiaptotropic substances). It was
not possible to inhibit their effect with ammonia. Euler,
Liljestrand and Zotterman therefore assumed that these subst-

ances, ivliich evidently do not act in the same -(vay as hypoxemia
and carbon dioxide, have tlieir point of attack on some synaptic
formation in the carotid bodies. Tlie morphological basis seems
to exist, inasmuch as it has been possible to demonstrate the
presence of cells in the carotid body partaking of the nature of
ganglion cells (De Castro; Meijlinc 1938). If there is a synapse
here, higher concentrations of nicotine, according to Langley,
would paralyse this. It also proved, in the experiments carried
out by Euler et al. that in larger doses both lobeline and
nicotine v’ere able entirely to inhibit all respiratory reflexes from
the sinuses. The chemical potentials disappeared entirely, while
the pressor potentials still came. There can thus in this case not
have been any paralysing effect upon the actual nerve fibres, hut
tlie observed effect must, in the light of what is so far known, be
located to some peripheral synapse or analogous formation.
Larger doses of lobeline (1.5 mg) in the carotid artery caused a
complete disappearance of all chemical action potentials, but
also a reduction of the pressor potentials arose. The same was
the case with potassium chloride in larger doses, which had a
paralysing effect upon both the chemoreceptors and the baro-
ceptors (Euler 1938). Curare also had a similar effect (Euler
1940) ill this respect, inasmuch as there was an almost complete
paralysis of the small impulses hut not of the big ones after an
intrasinusal injection of 0.25 ml of a 0.2 % curarine solution. On the
strength of these and other experiments Euler, Liljestrand and
ZoTTERMAN therefore came to the conclusion that hypoxemia,
carbon dioxide and cyanide take effect on the chemoreceptors
m the carotid body, while nicotine, lobeline, acetylcholine and
potassium ions exercise their effect somewhat more centrally upon
some synapse or analogous formation. In addition to this,
ZoTTEUMAN (1944) advances the suggestion that the specific action
of the lobeline may be located to the junction between the
afferent fibres and the chemo-sensitive cells of the carotid body,
thus affecting the excitatory transmission between receptor and
nerve fibre.

1. Technique and procedure.

Experiments have been performed upon cats which were anaes-
thetized with 0.O3 — 0.07 g of chloralose per kilogram of body-
weight. In order to obtain as much space as possible for the
electrodes on recording from the sinus nerve, the sinus region
was exposed hy severing of musculus sternocleidomastoideus and
hy removing the large lymph gland that covers the sinus region.
Tlie glossopharyngeal nerve is easily found where it passes along
the bulla tympani. After resection of the lateral part of muscidus
digastricus the glossopharj’iigeal nerve was severed as centrally
as possible. By gripping the central end of this nerve it is now
possible to dissect out the branch that rims to the carotid sinus.
In order, as usual, to obtain as high a signal-to-noise ratio as
possible, the sheath of the nerve was drawn off, care being taken
to see that the nerve was not exposed to any undue stretching or
squeezing. The depressor nerve was dissected out in the way
already described.

Into the central part of the lingual artery, which proceeds from
the external carotid artery a centimeter or so cranially to the
carotid sinus, and which runs along the hypoglossal nerve, a
cannula was introduced. After clamping the external carotid
artery above the point of departure of the lingual artery on the
occasion when the injection was to be made, this cannula enabled
a direct intrasinusal injection. A catheter was introduced in the
usual way through the common carotid artery and aorta, so that
the tip came to lie just inside the aortic valves.

The arterial blood pressure was recorded from the femoral
artery by means of an Hg-manometer. Intravenous injections
were administered through a cannula in the femoral vein.

2. Carotid body.

If on an intrasinusal injection of some few pg of lobeline
hydrochloride . one obtained the ordinary, tremendously strong
increase of the chemical impulses, one might be sure that the
chemoreceptors functioned. Lobeline in doses varying between

[/VV\A/WW\7W^

[aAAAA/W^AAAj'Wu^

Fig. 16. Cnl 3.2 kg. Spontaiicows rcfpirnfioti of air. Kcrording from tlic sinus
ncnc. A. Control, BP 130 nun Hg. B. After intrnsinusal injection of 0,5 mg
uretlianc in 0.20 ml Ringer's solution, BP 130 mm Tip.

25—50 jtg inlraeimifally caused, ns is known, n temporary reduc-
tion or complete extinction of tlie chemical impulses, while doses
of ahoul 0.25 mg generally caused a lasting extinction of both
the chemical impulses and the pressor impulses.

On intravenous injection of urethane to narcotize animals, one
may observe in connection with a momentary increase in the
rate of injection a temporary reduction or standstill of . the
respiration. In investigation of flic effect of urethane on
intrasinusal injection it proved that this substance, in doses of
about 0.25 — 0.5 mg, caused for a minute or so a disappearance of the
chemical impulses (Fig. 10). Tliis, ns well as, witliout doubt,
also a central effect, may provide the explanation of the effect
on the respiration that appears in connection with a more rapid
injection of urethane.

With 0,5 N acetic acid it was also possible to obtain an extinc-
tion of the chemical impulses. If one injected O.io — 0.i5 ml of this
acid intrasinusnlly one obtained in the majority of cases an
inhibition of the chemical impulses that proved permanent, or

Fig. 17 a. Cat 2.7 kg. Artificial respiration. The action potentials recorded
from the sinus nerve. A. Slight over-ventilation with air. Bloodpressure
120 mm Hg. B. Same ventilation but with 10.5 % oxygen in nitrogen. BP
130 mm Hg. C. After intravenous injection of 1 mg lobeline hydrochloride.
BP 160 mm Hg. Time: 1/50 sec.

that was at least not changed for several hours. The large pressor
impulses remained apparently unchanged. If one injected rapidly
a quantity tvhich was two or three times larger than tliat given
above, one obtained in the majority of cases also a disappearance
or considerable reduction of the pressor impulses. These, however,
returned successively in the course of five to ten minutes, while
the chemical impulses had definitively disappeared. On rapid
intrasinusal injection of O.o — 0.75 ml of this acid a lasting dis-
appearance of both the chemical impulses and the pressor

Fig. 17 l>. Same, cxperttncnt n.« in ftp. 17 n iiui oflcr ^nt^fl^intI^a^ injection
of 0.10 ml 0.5 Is' occiic ncid, 1). Sliplil ovcr-ventiliitinn ^*illi nir, BP 110 inni Hp.

E, Same ventilation But wiili 10.5 % oxvpcn in nitropen. BP 115 niin Up.

F. After intravenous injection of 1 nip lolieline liydroehloriilc, BP 130 nint Up.
Time: 1/SO see.

impulses vvns obtnined. As cnicrpes front this, there is a rather
big difference heltvccn the nniount required to eliminate only
the chemorcceplors and the nniount that definitively destroys
both clienio- and baroceptors.

Fig. 17b shows the effect of O.io ml of acid intrasinusally. In
comparison with the control (Fig. 17a) one secs here bow the
hypoxemia no longer provokes the ordinary increase of the
chemical impulse.s. This is also the case after iutravenotis injection
of 1 mg of lobcline hydrochloride, which before the elimination

C , , - .

Fip. 18. Cal 2.9 kg. Artificial respiration air. Recording of the action potent-
ials from the right depressor nerve. A. Control. BP 90 mm Hg. B. After injection
via a catheter with the tip inside the aortic valves of 0.10 mg lobeline
hydrochloride. BP 110 mm Hg. Thereafter elimination of the chemoreceptors
with 0.30 ml 0.5 N acetic acid. C. Same injection of lobeline hydrochloride
ns in B. BP 105 mm Hg.

of the chemoreceptors caused the ordinary increase of the
chemical impulses lasting for several minutes. Now one obtains
only a considerable increase of the pressor impulses, which is
connected "with tlie increased arterial blood pressure. There can
thus in this case not have been any paralysing effect on the nerve
itself, but only the chemoreceptors as tvell as the eventual sjuapse
formation (according to EuUEn, Liljestrand and Zotterman)
have been selectively eliminated.

L_l_l

Fig. 19. Cat 2.1 kg. After bilateral sinus denervation. Upper curve blood
pressure, lower Curve respiration recorded with u pneumograpb. Tbe first
arrow indicates intracardial injection of 0.3 nig lobcline bydrocliloride.
At the second arrow tbe cbcnioreceptors of ibc aorta are eliminated with
0.30 ml acid. The third and fourth arrows indicate tbe jioint of time for
intracardial injection of 0.3 and 0.5 mg respectively of lobelinc bydrocliloride.

3. Aortic body.

Analogous conditions, as %vas expected, could be demonstrated
from the aortic body on intracardial injection through the catheter,
whose tip was placed inside the aortic valves. Fig. 18 shows the
effect of 0.1 mg of lobeline hydrochloride before (B) and after
(C) the injection of O.ao ml of acid. Nor was this effect of the
acid reversihle for several hours, i. e. for the period during which
the effect ivas checked. Fig. 19 shows, in an animal with denerv-
ated sinuses, the effect on the respiration and blood pressure of an
intracardial injection of lobeline before and after the elimination
of the chemoreceptors of the aorta.

With this method one thus has a possibility of selectively
eliminating the chemical, reflexogenic excitability Avhile the
pressor regulation is unaffected.

YIl. The distribution of the effect of specific
stimuli between the carotid and

aortic bodies f

As all earlier experiments have sho^vn, the reflexogenic stimulus • •
from the chemoreceptors in the carotid bodies exercises the main
effect, while those from the aortic body are of more subordinate
importance.

In his investigation, CoMROE (1939) gives a scheme of this
percentual distribution between the carotid bodies and the aortic '
body in the dog of the effect of hypoxemia on the respiration
and blood pressure in connection with denervation experiments, ;
He finds that the carotid bodies alone are able to produce the 5
same stimulation upon the respiration as %vhen the animal is
intact, while when only tlie aortic body is functioning one only
obtains an increase of 25 % in the respiration, whereas the rise •-
in the blood pressure is now more clearly marked than in the
intact animal, owing to the fact that the regulators of the blood ^
pressure from the sinuses have been eliminated by the den- •-
ervation. In estimating the distribution of the effect of liyp- j
oxemia on the blood pressure and respiration between the chemo- ,
sensitive zones in the intact animal he finds that about 15 % ,

of the rise in the blood pressure is due to the reflexes from the
sinuses, while about 85 % comes from the reflexes of the aorta
region. The effect on tlie respiration is so distributed that 90 %
is localized to the carotid bodies, while only 10 % goes over
the aortic body.

1. Technique and Proeedure;

In the investigation of this distribution of the effect the chemo-
receptors have as usual been eliminated with 0.5 N acetic acid.
The experiments were performed, as before, on cats, \s'hich

.5fi

^\’ere first narcotized with ether and afterwards received O.or, —
O.OT g of chloralose per kilogram of hody-weight. A tracheal
cannula was introduced in the usual way. The blood pressure
was registered from the femoral artery with an Hg-manometer.
Cannulae were introduced into the lingual arteries and a catheter
was passed through the common carotid artery to the heart in
the way already described.

The respiration was recorded according to the principle
described by Haldane and Priestly (1905) with a method
described by Euler and Liljestrand (1936), the animal being
placed in an air-tight box ^vhich was connected with a spirometer.
The tracheal cannula was connected with a glass tube which
left the box through an air-tight opening in one wall and ■was
connected with a Muller valve with slight resistance, some milli-
meters of water. The ventilation was obtained as a product of the
spirometer reading and the respiratory frequency. The arterial
cannula passed through a hole in the w’all, enabling simultaneous
registration of the blood pressure.

2. Hypoxemia.

When the animals experimented upon were allowed for 2
minutes to breathe spontaneously a gas mixture poor in oxygen,
7.3 % oxygen in nitrogen, before and after the elimination in
turn of the chemoreceptors in the aorta region and in the sinuses,
the following results were obtained in nine typical experiments,
a) The respiration. The stimulating effect on the respiration
of the intact animal amounted to about 54 % as compared wilb
the ventilation with air. If the chemoreceptors in the aortic body
were eliminated one still obtained a rise in the respiration of
43 % on respiration of the gas mixture poor in oxygen. If after this
one destroyed also the chemoreceptors in the sinuses, the usual
reduction and standstill of the respiration appeared (Fig. 20)
that is obtained with hypoxemia after severance of the sinus
nerves and the vago-depressors. If air is not then quickly supplied
the cat dies.

Fig. 20. Cat 3.1 kg. Upper curve recording of respiration by means of a hodj
plcthysmograph, lower curv’e blood pressure. A. Spontaneous respiration of
7.3 % oxygen in nitrogen. B. After elimination of the chcmoreceptors in
the aortic body. Spontaneous respiration of 7.3 % oxygen in nitrogen. C, After
elimination also of the chenioreccptors in the carotid bodies. Spontaneous
respiration of 7.3 % oxygen in nitrogen.

If the cliemoreceptors in the sinuses were destroyed first one
obtained no increase in the respiration with hypoxemia tliat could
he ascribed to the aortic body, but the ordinary reduction and
standstill of the respiration appeared. From this it emerges that
the chcmoreceptors of the aorta alone cannot protect the cat in
case of hypoxemia.

h) The blood pressure. If we now consider the blood pressure
reactions during these experiments we find that tlie slight rise
in blood pressure that appears in the intact cat on hypoxemia
does not occur or is changed to a slight fall after elimination of
the chcmoreceptors of the aorta. If both these and the chemo-
receptors of the carotid bodies are destroyed, hypoxemia causes,
in the majority of cases, an initial rise in the blood pressure.

whicli occurs already l)efore the respiration has been changed,
hut which is tJiercafter followed hy a marked fall in the blood
pressure (Fig. 20). This initial rise must, to begin with, he consid-
ered to be due to a direct effect of the poorly oxygenated blood on
tlie vaso-motor centres. But the rise in blood pressure also remains
for a while after the respiration has diminshed or slopped, so
that one must also consider an effect of an increase of carbon
dioxide on the centres. The marked subsequent fall in the blood
pressure is a sign that these centres and the heart arc failing.

Selladurai and Wriciit (1932 a) also observed in some cases
this initial rise in blood pressure before the appearance of the
definitive fall in blood pressnre. In other cases they obtained
only the fall in blood pressure, without any preceding rise, in
experiments on cats. This has also been observed in the investiga-
tions referred to above. Euler and Liljestkand (1936), in
experiments on cats, obtained only the fall in blood pressure,
whereas on the other hand, they found in dogs the initial rise
before the respiration had changed, but follo^v•ed thereafter by
tlie usual fall in blood pressure. They therefore assumed that
tlie vaso-motor centre in cats was especially sensitive to hypoxemia
and that it accordingly soon failed.

3. Lobclinc.

a) I'he respiration. The quantitative distribuVimr of the effect
of lobeline on the respiration has been studied before and after
tlie destruction of the chemoreceptors. Lobeline hydrochloride
has been given intravenously in an amount of O.o mg, wdiich gave
rise to a suitable marked increase of the respiration. This
increase of tljc ventilation amounted in the intact animal in nine
typical experiments to 41 % as compared with the original value.
After elimination of the receptors in the aortic body the stimulat-
ing effect of lobeline on the respiration diminished to 30 %.
If after this the chemoreceptors in the sinuses were destroyed, no
effect on the respiration was obtained (Fig. 21). If the chemo-

Fip. 21. Cat 3.3 kg. Upper curve recording of respiration witli a liody pleth-
yainoprapli, lower curve blood pressure. At tbe arrows intravenous injection
of 0.9 nip lobelinc liydroebloridc. A. The effect on the intact animal. B. After
elimination of the chcmoreccptors of the aorta. C. After elimination of the
chcmoreccptors in the sinus.

receptors were eliminated in the reverse order the result was
the same.

b) The blood pressure. Apart from the stimulating effect on
the respiration, loheline also has such effect on the blood pres-
sure. This effect has been mainly interpreted as central, although
a part of it is undoubtedly to be considered as caused by a stimulus
from chemoreceptors in the carotid bodies (Hey.mans, Bouckaeut,
Euler and Dautrebande 1932, Euler, Liljestrand and ZoT-
TERMAN 1939) and in the aortic body (CoMROE 1939). Ligation of
the suprarenal bodies did not prevent the rise in blood pressure.

If one compares the rise in blood pressure caused by lobeline
in the intact animal with the rise that is obtained after the
receptors in the glomus aorticum have been destroyed, one gets,
in the majority of cases, a slight reduction. Subsequent elimina-
tion of the chemoreceptors in the sinuses caused in most cases
a further reduction in the rise in blood pressure, but the results
are so varying that one cannot give any exact percentual reduction.

4. Hypercapnia.

In this connection the effect of carbon dioxide on the respira-
tion and blood pressure befon; and after the selective elimination
of the chemoreceptors was also investigated.

a) The respiration. In order to ascertain to what extent the
eventual respiration-stimulating effect of carbon dioxide was
conditioned by stimulation of the chemoreceptors, the animals
were allowed to breathe spontaneously a gas mixture of 7.3 %
of carbon dioxide in air before and after the chemoreceptors had
been eliminated. If the aortic body was destroyed one obtained
a reduction that was so insignificant that it could not be estimated
with any certainty. On the other hand, one obtained a reduction
of about 16 % if the receptors in the carotid bodies tvere
destroyed. Bearing in mind that also the alveolar carbon dioxide
tension had probably increased, and thus caused a stronger central
effect after elimination of the chemoreceptors (Euler and
Liljestrand 1936), one is probably well advised to consider this
figure as possibly somewhat too low.

b) The blood pressure. On inhalation of the carbon dioxide-rich
gas mixture the blood pressure was either unchanged or shotved
ah extremely slight rise. No certain change after destruction of
the chemoreceptors was obtained.

VIIL The reflex efiect on the respiration of
intrasinusal pressure changes

It has long been known that a compression of tlie common
carotid arteries causes an increase in the respiration. To begin
with, this was considered to be due to an impaired central blood
supply. The discovery of the aortic and sinus reflexes led to a
closer investigation of this phenomenon. It was then found that
a rise in pressure in the sinus region caused an inhibiting effect
on the respiration, while a reduction of the pressure in the same
region, e. g. by compressing the carotids, gave rise to an increase
in the respiration (Hevmans and Bouckaekt 1930, Koch and
Maek 1931, Goluyitzer-Meier and Schulte 1931, Schmidt 1932,
Heymans, Bouckaert and Dautrebande 1932). Gemmill,
Overstreet and Hellman (1933), however, did not find any
change in the respiration in non-anaesthetized dogs on compres-
sion of the carotids, either before or after sinus denervation.
Heymans and Heymans (1927) were tlie first to show the inhibit-
ing effect on the respiration of the rise in pressure in the reflex-
ogenic zone of the aorta.

Heymans and Bouckaert asserted that the respiratory centre
was subject to a reflexogenic influence from the aorta and sinus
regions, and they were able to sho^v, in perfusion experiments
on the sinus, tliat the respiration could be reduced and even
stopped if the perfusion pressure tvas raised sufficiently, whereas
a low perfusion pressure increased the respiration. These effects
did not appear after bilateral sinus denervation. Nor was it
possible to obtain the usual increase in tlie respiration after
compression of the common carotid arteries if the sinus nen'es
were severed. Tliey considered that this effect on the respiration
was produced from the baroceptors. Schmidt (1932) assumed
that the effect on the respiration of blood pressure changes in

the sinus region was presumably due to other factors than those
which affect the circulation, as the respective intensities of the
two phenomena were often completely independent of each
other. This assumption has received support in several investiga-
tions. Euler and Liljestrand (1936, 1937) considered that the
changes of the circulation in tliis region caused by variations in
the blood pressure changed the chemical effect of the blood. They
came to the conclusion, therefore, that the reflexogenic effect
on tlie respiration in connection -with various intrasinusal pres-
sures was not elicited from the baroceptors but from the chemo-
receptors, owing to a change in the chemical effect of the blood.
Thus, low intrasinusal pressure leads to hypoxemia and hyper-
pnoea, high intrasinusal pressure gives rise to improved circulation
and a disappearance of the hypoxemic and carbon dioxide
stimuli, which quickly leads to a reduction of the ventilation
and frequently to apnoea until the carbon dioxide losses have
been made good. Euler, Liljestrand and Zotterman (1939),
recording tlie action potentials from the sinus nerve of the cat,
were able to show a considerable increase of the chemical impulses
on compressing the carotids; but this increase of impulses dis-
appeared if tliey over-ventilated the animal tvith oxygen, and
must therefore be considered to be due to stimulation of the
chemoreceptors. Rudberg (1938) showed that the stimulating
effect on the respiration followng compression of the carotids
was increased if the blood pressure was reduced below a critical
value. The same author (1940) showed that this could be counter-
acted to a certain extent by inhalation of oxygen. Further
investigations by Euler and Liljestrand (1940) showed that on
perfusion of the sinus region with blood poor in oxygen (saturation
= 33 %), in contradistinction to blood rich in oxygen, a reduc-
tion of the intrasinusal pressure had no effect on the respiration,
as in this case the chemoreceptors, presumably already at high
perfusion pressure, were maximally stimulated by the hypoxemia.

Winder (1938), by embolization of the carotid body tvith a
lycopodium suspension, attempted to eliminate the chemo-
receptors, but respiratory reflexes still appeared, however, on
intrasinusal pressure variations. Marri and Hauss (1939) carried

out an investigation on the effect of certain factors on the resp-
iratory reflexes in connection with intrasinusal pressure variations
employing a method which Avas previously used hy Hetmans,
Don'ATELei and Shen (1938). A rubber bag or a bag made out of a
vein Avas introduced into the sinus, and they considered that by this
means, by varying the pressure in the bag, they could release
pressor reflexes Avhithout affecting the chemoreceptors. By
increasing or reducing the pressure they Avere able to affect the
respiration. This Avork, as Avell as a later one by Grimson and
Shen (1939), in Avbich tlie same method Avas employed, has been
exposed to a certain criticism, as Avith this procedure the blood
supply to the carotid body must have been affected.

Bjurstedt and Hesser (1942) used a special cannula, thanks
to Av’hich the sinus regions could be relatively Avell supplied Avitli
oxygenated blood ev'en at a Ioav perfusion pressure. In tbis con-
nection they Avere able to shoAV tbat endosinusal pressure changes
did not necessarily cause any change in the respiration even if
typical blood pressure reflexes appeared. Under certain circumst-
ances the change in pressure could result in a variation in the
respiration, even if no blood pressure reflexes appeared. They
therefore considered that the respiratory changes that arise in
connection Avith intrasinusal pressure changes are released from
the chemoreceptors. Bjurstedt and Euler (1942) also found
that under normal conditions the respiration Avas not affected
reflexogenically from the baroceptors in tbe sinus and aorta
regions, altbough they Avere able definitely to confirm the
occurence of such an influence after severance of the vago-
depressors.

To the method of selectively eliminating the chemoreceptors
AAdth acid it might be objected tbat also the baroceptors are
damaged. That this is scarcely the case, however, has already
been pointed out in connection Avith recording of the action
potentials from tlie sinus and depressor nerA'es. Further support
for the assumption that they are intact is given in the folloAving
experiments Avith simultaneous recording of the respiration and
blood pressure, in Avhich, as a test of the intactness of the pressor-
reflexes after elimination of the chemoreceptors, the effect on

\ .■m.v.'iS'Jjl

'’tT«»p“*V°otd« '»

aastaawo ““'a'trfter '! jl a cV.»»6e

He*'* cetleaase"' '“'1 joioel taji « ” eteceP'"”’. "

on ibat 1 jieisate "> ,c tlje c jirauoa

' ‘'» ”n“”X.e!'.'d '• ''’“’'^\re t»oteeeV“”

'""n ir' ena aid “^""atlety

proved VhAt catoua

..ere se\ecUve\y

IX. Adrenaline apnoea

If one injects a sufficiently large amount of adrenaline intraven-
ously, it proves that the ordinary rise in bipod pressure is
followed by a reduction of the respiration, which is sometimes
so pronounced that apnoea occurs. This phenomenon was first
described by Oliver and Schafer (1895), wdio injected extract
of the supra-renal body into dog and rabbit. A similar effect on
the respiration was also obtainable if one caused a quick rise
in blood pressure by compressing aorta abdominalis (Roberts
1921). He explained that the adrenaline effect on tlie respiration
was caused by a direct effect on tbe medulla through a constric-
tion of the vessels in this region. This led to a paralysis of the
centre through hypoxemia or accumulation of metabolites. Mel-
LANBY and Huggert (1923) also arrived at tliis view. This
explanation, however, is not easily reconciled with the apnoea
which can be provoked on compression of the aorta. Since the
sinus mechanism has been discovered new ways of explaining the
adrenaline apnoea have been sought. Hetmans and Bouckaert
(1930) and Wright (1930) sho^ved that it is mainly reflex-
ogenically conditioned, for after denervation of the carotid sinuses
and severance of the vagi adrenaline caused practically no, or an
extremely slight, effect on the respiration. Heymans and
BoucKj\ert interpreted the phenomenon to mean that the
adrenaline apnoea is due to a reflexogenic inhibition of the
respiratory centre provoked by tbe rise in blood pressure in the
aorta and sinus regions: thus a stimulation of the baroceptors.
The earlier cited literature on the reflex effect of intrasinusal
pressure changes on the respiration tends rather to support an
explanation of the adrenaline apnoea as being due to the increased

Fig. 23. Cal 3.2 kg. Recording of llic action potentials from the right depressor
nen'e. Artificial respiration nir. A. Control, BP 140 mm Hg. B. After
intravenous injection of IS pg adrenaline, BP 200 mm Hg.

arterial blood pressure, which causes a better circulation through
the carotid and aortic bodies and thus a reduction of the effect
of the chenioreceptors in these regions. Gernandt, Liljestrand
and ZoTTEil.MAN (1945) gave further evidence that the adrenaline
apnoea was caused by a reduction in the activity of the cheino-
receptors. In experiments on cats they showed that the apnoea
cotdd be prolonged if the animal was allowed to breathe oxygen,
and more or less suppressed if instead it was allowed to breathe
oxygen-poor or carbon dioxide-rich gas mixtures. Records of the
action potentials from Hering’s nerve showed that the chemical
impulses were strongly reduced during the adrenaline apnoea.
Taking these last investigations as our point of departure, it
thus appears rather doubtful that the adrenaline apnoea should
be caused by a directly inhibiting effect from the baroceptors
in the reflexogenic sinus and aorta regions.

Fig. 24. Cat ."^.l kg. Same experiment as in fig. 22. Upper curve recording of
respiration ivith a body pletbysmograph, lower curve Idood pressure. The
arrows indicate intravenous injection of 20 pg adrenaline. Between A and B
the cliemoreccptors in the aorta and sinuses are eliminated.

1. The effect of intravenous injection
of adrenaline on the activity in the chcmoceptivc
fibres in the depressor nerve.

On recording the action potentials from the depressor nerve,
the same considerable reduction of the small impulses after
intravenous injection of adrenaline (Fig. 23) was obtained. As
has already been pointed out, it is technically impossible to
destroy with mechanical means the pressoceptive fibres without

at the same time injuring the cliemoceptive fibres. The pressor
impulses thus remain, and one sees how they increase quite
naturally after the injection of adrenaline.

2. The effect on the respiration of intravenous adrenaline
injection after elimination of the chemoreceptors.

According to the previously desbrihed method, the sensitivity
of the chemoreceptors was extinguished with acid, after which
it was attempted to ascertain whether adrenaline can then cause
any apnoea. It proved that after the above procedure one could
not obtain any adrenaline apnoea on intravenous injection of
20 pg of adrenaline, which otherwise caused a clear apnoea (Fig.
24). Compression of the carotid artery after spontaneous breath-
ing of pure oxygen for 2 minutes gave rise to the usual rise in
blood pressure. Sometimes, however, one still obtained a slight
remaining adrenaline effect on the respiration after the destruc-
tion of the chemoreceptors. Heymans and Bouckaert, Wright
and Gernandt, Liljestrand and Zotterman also found this
effect after severance of the sinus nerves and the vago-depressors.
The last-mentioned authors also observed that this slight adren-
aline effect on the respiration remained during the inhalation
of oxygen. It would seem most natural to assume that it is caused
by a central effect, although the actual mechanism is unknown.

These results give further support for the vie\v that the
adrenaline apnoea is caused by a reduction or disappearance of
the reflexogenic, chemical stimulation of the respiration, and
not by any inhibiting effect exerted on the centre from the baro-
ceptors.

X. Tonic chcmoreflex respiratory
stimulation

A reduced respiration after the elimination of the sinus
mechanism is a phenomenon that has been observed by several
authors in experiments on animals. Selladurai and Wright
(1932 b), in experiments on cats narcotized with chloralose found
that sinus denervation caused a reduction of the ventilation
amounting as an average to 22 %. In experiments on decerebrated
cats they obtained on denervation a reduction of 33 %, doubtless
partly owing to the fact that the carotids were ligated and the
respiration was in consequence reflexly increased before the
denervation. Stella (1936) found in anaesthetized dogs a
similarly reduced respiration on fimctional elimination of Her-
ing’s nerves. Also Euler and Liljestrand (1936, 1940) observed
this reduction in the ventilation and an increase in the alveolar
carbon dioxide tension on severing the sinus nerves and the
vago-depressors in experiments on dogs and cats. The effect was
as a rule reduced but not eliminated if the animals "were allowed
to breathe pure oxygen. Gesell, Lapidez and Levine (1940) and
Schmidt, Comroe and Dripps (1939) also foimd this reduction
in the respiration after elimination of the sinus mechanism. These
experiments thus show that under the conditions obtained there
is a flow of impulses in the sinus nerves which affects the ventil-
ation. This flow of impulses is elicited by hypoxemia and carbon
dioxide according to the records of tlie action potentials from the
sinus nerve.

To obtain a quantitative idea of the magnitude of this tonic
stimulation from the chemoreceptors on the respiration in these
experiments is very difficult, since .severance of the sinus nerve
breaks the afferent channel for the chemoreceptors as well as for
the baroceptors: and, as has been mentioned above, several authors

consider that from tlie baroceptors there is produced a tonic,
refJexogenic inhibiting effect on the respiration in agreement with
the mhibition exerted on the blood pressure (Heymans and
Bouckaeut 1930, Heymans, Bouckaert and Dautrebande 1930,
Koch and Mark 1931). These writers found that after bilateral
sinus denervation and after severing both vago-depressors
an increased respiration was obtained. Elimination of the
tonic inhibition of the respiration would thus have a
directly opposite effect on the absence of the tonic influence
from the chemoreceptors. Tlic reason for this difference
of opinion is perhaps to seek in tlie change in the respiration
that arises when the respiration-regulating function of vagus is
eliminated. That the chemical factor is in any case far more
important has been shown by Berntahl and Weeks (1939), wdio
cooled the chemoreceptors in connection with isolated circulation
of these with blood at different temperatures. In this way they
ivere able to obtain a selective elimination of these ^vhiIe the
pressor reflexes were intact: and in experiments on dogs they
found that an elimination of the tonic, reflexogenically stimulat-
ing effect from the chemoreceptors entailed a reduction of tlie
ventilation amounting to 34 %.

The method of selectively eliminating the chemoreceptors in
both the sinus and aorta regions rvith acid offers fresh possibilities
of studying the tonic, reflexogenically stimulating effect on the
respiration and blood pressure released from all the chemo-
receptors that have any physiological importance. The pressor-
reflexes can remain intact and function as before.

The animal w'as placed in the above-described body plethysmo-
graph. The registration of the respiration was begun about an
hour after the animal had been narcotized with ether and
chloralose. It was therefore not necessary to reckon with any ether
effect on the respiration. As Geunandt (1943) pointed out also
in experiments on cats, with this metliod of recording one obtains
during the first hour a continuous successive reduction of the
respiration, after which it remains relatively constant. In view
of this fact the ventilation has not been estimated during this
first hour.

Fig. 25. Cal 3.1 kg. Registration of tlic respiration hy means of a body pletli-
ysniograpli. A. Spontaneous respiration of air. B. After selective elimination
of the cliemoreceptors in aorta and sinuses.

Before and after elimination of the cliemoreceptors the effect
of inhalation of 7.3 % oxygen in nitrogen ivas tested.

1. The effect on the respiration.

The magnitude of the ventilation was estimated in periods of
1 minute. The mean value in seven typical experiments on cals
weighing helween 2.5 — .3.4 kg was a ventilation of O.sn .+ O.os liters
per minute. The respiratory frequency was ll.o ."t. O.tg hreaths
per minute.

If after this the cliemoreceptors in the aorta region were elimin-
ated, the ventilation was reduced to O.in .+. O.02 liter per minute,
with a respiratory frequency of IO.g +- O.gi hreaths per minute.
A selective elimination of the cliemoreceptors in the sinuses caused
a furllier reduction of the respiration to a value of O.30 :r- O.02
liters per minute. The respiratory frequency was now 8.g .i. 6.02
breaths per minute (Fig. 25). This thus shows that the ventilation
is reduced by about 36 % if the reflexogenically stimulating effect
from the cliemoreceptors under these conditions is eliminated.
If one considers the distribution of this effect between the chemo-
rcceptors in the aortic body and the carotid bodies respectively

one arrives at the result that out of this 36 %, 13 is conditioned
over the former while 23 is produced by the latter.

In view of tlie fact that the circulation in the right common
carotid artery was cut off by the catheter and the sinus on this
side did not obtain blood through the ordinary channels, elimin-
ation of the chemoreceptors was also effected in tlie reverse order.
The chemoreceptors in the carotid bodies were first eliminated,
and after this the catheter was introduced into the common
carotid artery and acid was injected to destroy the chemoreceptors
of the aorta. It was not possible, however, to observe any differ-
ence in the results obtained. Nor was any definite change in the
results given above obtained after severing tbe right sinus nerve
with subsequent elimination of the functioning chemoreceptors.
This is in agreement with the observation tliat denervation of the
one sinus scarcely causes any change, as the other sinus exercises
a compensatorily increased function.

2. The effect on the blood pressure.

The elimination of the chemoreceptors at the moment of
injection caused a very brief rise in blood pressure, which there-
after returned to the original level. This rise in the blood pres-
sure is probably to be considered as mainly due to a general
sensitive stimulating effect of the acid, as one sees with what
strong jerks the animal reacts. The disappearance of the pressor
impulses for some moments that is sometimes observed w'hen
recording tlie action potentials may also be a contributing cause
through the temporary absence of the pressure regulation.
Theoretically, the elimination should entail a lowering of the
blood pressure through the removal of the tonic stimulation of
the chemoreceptors with effect on the blood pressure. But apart
from this temporary rise in blood pressure the pressure remained
constant, if one excepts the usual gradual lowering that is
obtained in experiments that are continued for several hours.
This is of course also to be expected, as the pressor regulation
is intact. The reduction in the ventilation due to elimination of

Fig. 26. Cat 3.0 kg. Upper curve recording of respiration with a body plcth-
ysmograpli, Jower curve blood pressure. A. Bet^veen the arrou's clamping of
the carotid artery after 2 minutes’ spontaneous respiration of pure oxygen.
B. The cliemorcceptors in the sinus and aorta regions selectively eliminated
with acid. Between the arrows clamping of the carotid artery after 2 minutes'
spontaneous respiration of pure oxygen. C. After further injection of acid
elimination also of the baroceptors. Between the arrows clomping of the
carotid artery after 2 minutes’ spontaneous respiration of pure oxygen.

the chemoreceptors is also followed, as Euler and Liuestrand
(1936) have shown in connection with denervation experiments,
by an increase in the alveolar carbon dioxide tension. This might
possibly be able to cause a rise in the blood pressure; but one
must, however, also bear in mind that a certain hypoxemia may
arise, which in animals with eliminated chemoreceptors causes
a lowering of the blood pressure. It is possible that these two
factors compensate each other, so that the blood pressure remains
constant.

In this connection it also proved suitable to return to the
question of the importance of the baroceptors for the regulation
of the respiration. At first the chemoreceptors were destroyed
with acid, and the usual reduction of the ventilation appeared,
after which O.s ml acid was quickly injected into each sinus as
^vell as 0.75 ml via the catheter. This immediately caused a strong.

permanent rise in the blood pressure, •which sho’^ved that the
depressor reflexes tvere eliminated. A further possibility of
ascertaining whether the depressor reflexes were eliminated lay
in a compression of the carotid artery. Fig. 26 shows the effect
on the blood pressure and respiration of this compression, after
the animal had first been allowed to breathe 100 % oxygen
spontaneously for 2 minutes in order to exclude any central effect.
Fig. 26 A shows the effect of the compression in the intact
animal, B after the chemoreceptors have been destroyed, in C
also the baroceptors have been destroyed, and then, as usual,
one does not get this typical rise in blood pressure in connection
with the clamping of the artery. It was not possible to obtain
any increase in the ventilation indicating the elimination of any
tonic inhibition exerted from the baroceptors. The ventilation
in the intact animal before the compression "was 0.42 liters per
minute, with a respiratory frequency of 8.8 breaths per minute.
After destruction of the chemoreceptors the ventilation was as
usual reduced, and amounted then to 0.24 liters per minute, ^^’itll
a respiratory frequency of 7.8 breaths per minute. This ventilation
tvas not changed after the elimination of the baroceptors. These
results thus constitute direct evidence against the assumption
that the baroceptors exercise any tonic inhibiting effect on the
respiration of the cat.

XI. Summary

Taking previous investigations on the existence of a chemo-
sensitive region in the tract around aorta ascendens, this region
has been made the object of closer studies.

In an anatomical survey a more detailed account lias been
given of the localization of a number of cell groups of parag-
anglionic type around the aorta and the pulmonary artery. These
so-called paraganglia are in part sympathetic, consisting chiefly
of chromaffin cells, and in part parasympathetic, which have the
same morphological structure as the formations occurring in
the sinus region.

The chemo-sensitive zone of the aorta in the cat has been
localized to the region below the aortic arch in the connective
tissue between the aorta and the pulmonary artery. By mechanic-
ally destroying this formation as well as tlie afferent nerve fibres
proceeding therefrom, one obtained, on recording the action
potentials in the depressor nerve, the afferent nerve from the
aortic body, a complete elimination of all chemical impulses,
while impulses from the baroceptors still continued to come.
Owing to the anatomical conditions it was not possible in this
way to destroy only the afferent fibres of the baroceptors,
mthoul tlie afferent fibres of the chemoreceptors being at the
same time more or less involved in the process.

By tlie local application of certain specific stimuli it was
possible to confirm that the aortic body in the cat obtains its
blood supply through the coronary vessels. This local application
also enabled one to obviate the possibility of other paraganglia
around the aorta refle.xly affecting the respiration.

Action potentials in the depressor nerve released from the
chemoreceptors in the aortic body under the influence of
hypoxemia, carbon dioxide and specifically stimulating substances
(lobeline, piperidine etc.) have been recorded. These potentials

are of tlie same nature as those one obtains under similar condi-
tions on recording from the sinus nerve.

A method of eliminating the chcmoreceplors in the aortic body
and the carotid bodies by injecting locally a small amount of
0.5 N acetic acid is described. In this connection the chemo-
receptors are destroyed, while the baroceptors remain intact.
This fact u’as checked partly by recording the action potentials
from the afferent nerves and partly also by controlling the blood-
pressure reflexes on clamping the carotid artery. Larger amounts
of acid were also able to destroy the baroceptors.

After elimination of the chemoreceptors no effect of carotid
compression on the respiration was obtained. No sign that
the intact baroceptors in the cat can have any effect on tlie
respiration has been observed. As further support for this it was
possible to show that the ventilation did not undergo any change
when the baroceptors ^v'ere destroyed in an animal with chemo-
receptors already eliminated.

A selective elimination of the chemoreceptors caused a reduc-
tion in the ventilation by about 36 %. The chemoreceptors thus
exercise a tonic, reflex'ogenic stimulation of the respiration.
It was possible to measure the relative effects on respiration and
blood pressure which were exerted by hypoxemia and carbon
dioxide and such a specific stimulus as lobeline. This was done
after first eliminating the chemoreceptors of the aortic region and
then also eliminating those in the carotid sinus regions. This
separation of aortic and carotid sinus effects demonstrated the
subordinate importance of the aortic body as compared with the
carotid bodies.

References

Addison, W. H. F., and J. H. Comroe, Jr., Anat. Record. 1938. 70. supp. 3. 2.
Bernthal, T., and W. F. Weeks, Am. J. Physiol. 1939. 127 . 94.

Beyne, j., j. Gautrelet and N. Halpern, C. R. Soc. Biol. Paris. 1933. 113 . 585.
Bjurstedt, H., and U. S. v. Euler, Acta Physiol. Scand. 1942. 4. 23.
Bjurstedt, H., and C. M. Hesser, Ibidem 1942. 4 . 5.

Bouge, j. Y., and G. Stella, J. Physiol. 1934. 82 . 23 P.

Bouce, j. Y., and G. Stella, Ibidem 1935. 83 . 459.

Boyd, J. D., Pub. Carnegie Instit. of Washington 1937. 479 . 1.

Bronk, D. W., Proc. Soc. exp. Biol, and Med. N. Y. 1931. 28 . 1014.

Bronk, D. W., and G. Stella, Ibidem 1932 a. 29 . 443.

Bronk, D. W., and G. Stella, J. Cell. comp. Physiol. 1932 b. 1 . 113.
Bronk, D. W,, and G. Stella, Am. J. Physiol. 1934. 110 . 708.

Busachi, P., Arch. Ital. Anat. 1912. 11 . 352.

Clementi, a.. Boll. Soc. Biol. sper. 1928. 3 . 32.

Comroe, J. H. Jr., Am. J. Physiol. 1939. 127 . 176.

Comroe, J. H. Jr., and C. F. Schmidt, Ibidem 1938. 121 . 75.

Crimi, G., Boll. Soc. Biol. sper. 1933. 8 . 122.

Dautrebande, L., and R. Wecria, Arch. int. Physiol. 1937. 44 . 425.

De Castro, F., Trav. Lab. Recherch. biol. Univ. Madrid 1926. 24 . 365.

De Castro, F., Ibidem 1927. 25 . 331.

DrUner, L., Dtsch. med. Wschr. 1925. 51 . 559.

V. Euler, U. S., Skand. Arch. Physiol. 1938. 80 . 94.

V. Euler, U. S., Nordisk Medicin 1940. 7. 1389.

V. Euler, U. S., Acta Pharmacol. 1945. 1 . 29.

V. Euler, U. S., and G. Liljestrand, Skand. Arch. Physiol. 1936. 74 . 101.

V. Euler, U. S., and G. Liljestrand, Ibidem 1937. 76 . 27.

V. Euler, U. S., and G. Liljestrand, Acta Physiol. Scand. 1940. 1 . 93.

V. Euler, U. S., and G. Liljestrand, Ibidem 1943. 6. 319.

V. Euler, U. S., and Y. Zotterman, Ibidem 1942. 4 . 13.

V. Euler, U. S., G. Liljestrand and Y. Zotterman, Skand. Arch. Physiol.
1939. 83 . 132.

V. Euler, U. S., G. Liljestrand and Y. Zotterman, Acta Physiol. Scand.
1941. 1 . 383.

Gemmill, C. L., and D. L. Reeves, Am. J. Physiol. 1933. 105 . 487.

Gemmill, C. L., E. M. K. Geilinc and D. L. Reeves, Ibidem 1934. 109 . 709.
Gemmill, C. L., E. W. Overstreet and L. M. Hellman, Ibidem 1933. 104 . 443.
Geppert, j., Z. Klin. Med. 1889. 15 . 208.

Gernanit. B., Acln Phygiol. Scnnd, 1943. 6. 233.

Gernandt, B., and Y. Zottf.rm.vn, Ibidem 1945. 9, 112.

Gernandt, B., and Y. Zotteiiman, IJ»ideni 1945. 9. 362.

Gernandt, B., G. Liuesthand and Y. ZoTTi;nMAN, Ibidem 1945. 9. 367.
Geseel, R., and J, LAriDES, Proc. I6»b Ini. Pliysiol. Congr. (Ziirich) 1938.
I. 45.

Ge-seli, R., J. Lapides and M. Leain, Am. J. PJiysiol. 1940. 130. 155.

Gesele, R., and C. Moyer, Ibidem 1937. 119. .55.

Geseee, R., H. Krueger, G. Gorham and T. Rosenthai, Ibidem 1930. 40. 54.
Goeewitzer-Meier, Kt,., Pflugers Arch. 1934. 234. 342.

GoeewitzeR'Mkier, Kl., and E. Lerche, Ibidem 1940. 244. 145.
Goeeh’Iizer-Meier, Kl., and H. Schulte, Ibidem 1931, 229. 251.
Goorsuchtich, N., and R, pAN'NTf;n, C, R. Sor. Biol, Paris 1936. 121. 882.
Goormachtich, N., and R, Pan.mer, Iliidcm 1936. 123. 1261.

Grisison, K. S,, and T. C. R. Shen, Arcli. ini. Pbarmneodyn. 1939, 63. 95.
Haldane, ]. S., and J. G. Priestly, J, Physiol. 1905. 32. 225.

Hebinc, H. E., PnOgers Arcli. 1924. 206. 721.

Hetmans, C., Ycrli. Gcs. Krcisl. Forsrli. 1. Tagung. 1928. 92.

Het.aeans, C., and J, J. Bouckaert, J. Physiol. 1930. 69. 254.

Hetman.s, C., and J. J. Bouckaert, C. K. Soe. Biol. Pari.s 1933. 112. 1240.
Hey.mans, C., and J. J. Bouckaert, Ergebn. Physiol. 1939. 41. 28.

Heymans, C., j, j. Bouckaert, and L. Dautredande, Arcb. int. Pharmn-
cod)Ti. 1930. 39. 400.

Heymans, C., J. J. Bouckaert, and L. Dautrerande, liiidem 1931. 40. 54.
Hetmans, C,, J. J. Bouckaert and L. Dautrerande, Pfliigcrs Arcli. 1932.
230. 283,

Heymans, C., J. J. Bouckaert and P. Recniers, Le sinus caroiidien, Pnri.s
1933.

Heysians, 0., J, J, Bouckaert, U. S. v. Euler and L. Dautrerande, Arch,
int. Pharmacodyn. 1932. 43. 86.

Heymans, C., J. J. Bouckaert, S. Fauer and F. J. Hsu. Ihidem 1936. 54. 129.
Heymans, C., L. Don.atei.li and T, C. R. Shen, C. R. Soc. Biol., Paris 1938.
128. 784.

Heymans, C., and H. Handov.sky, Ibidem 1935. 119. 542.

Heyman.s, j, F., and C. Heyman.s, Arch, int. Pharmarodyn. 1927. 33. 273.
Heymans, C., and P. Rulant, C. R. Soc. Biol. Paris 1933. 113. 69.
Hollinshead, W. H., j. Comp. Neurol. 1939. 71. 417.

IWANOFF, G. H., Z, Anat. n. Entw. Gcsch. 1925. 77. 234.

Joncrloed, M. j., Ann, de Physiol. 1936. 12. 457.

Koch, E., and R. E, Mark, Z. Kreisl. Forsch. 1931. 23. 319.

Koester, G., and A. Tscher.mak, Arch, f, Anat, ii, Phy.siol., Anat. Alit. Suppl.
1902. 26. 255.

Kohn, a., Arcb. mikrosk, Anat. u. Entw. Gcsch. 1900. 56. 81.

Lambert, E., and E. Geli.hohn, Proc. Soc. Exper. Biol. N. Y. 1938. 38. 427.

Marri, R., and W. Hauss, Arch. int. Pharmacodyn. 1939. 63. 449.

Meijlinc, H. a., Bail und Innervation von Glomus caroticum und Sinus
caroticus. Utrecht 1938.

Mellanby, J., and A. Sx. G. Huccett, J. Physiol. 1923. 57. 395.

Moissejeff, E., Z. ges. exp. Med. 1927. 53. 696.

Muratori, G., Boll. Soc. Ital. Biol. Sper. 1934. 9. 1041.

Muratori, G., Monit. Zool. Ital. 1935. 45. 300.

Nonidez, j. F., Am. J. Anat. 1935. 57. 259.

Nonidez, j. F., j. Anat. 1936. 70. 215.

Nonidez, J. F., Anat. Record. 1937. 69. 299.

Oliver, G., and E. A. Schafer, J. Physiol. 1895. 18. 230.

Owen, H., and R. Gesell, Proc. Soc. Exp. Biol. N. Y. 1931. 28. 765.
Pacano, G., Arch. Ital. Biol. 1900. 33. 1.

Palme, F., Z. mikrosk. Anat. Forsch. 1934. 36. 391.

Penitschka, W., Med. klin. 1930. 26. 312.

Penitschka, W., Z. mikrosk. anat. Forsch. 1931. 24. 24.

Perman, E., Hjartnerverna hos hogre daggdjur och manniska. Stockholm 1920.
Rabl, H., Arch, mikrosk. Anat. 1922. 96. 315.

Riecele, L., Z. Anat. u. Entw. Gesch. 1928. 86. 42.

Roberts, Ff., J. Physiol. 1921. 55. 346.

Rudberc, T., Skand. Arch. Physiol. 1938. 79. 8.

Rudberg, T., Acta Physiol. Scand. 1940. 1. 89.

Saaman, a., and G. Stella, J. Physiol. 1935. 85. 309.

Schmidt, C. F., Am. J. Physiol. 1932. 102. 94.

ScHiMDT, C. F., and J. H. Comroe Jr., Physiol. Rev. 1940. 20. 115.

Schmidt, C. F., J. H. Comroe and R. D. Dripps, Proc. Soc. Exp. Biol. N. Y.
1939. 42. 31.

Schumacher, S., Anat. Anz. 1902. 21. 1.

Schweitzer, A., and S. Wright, Quart. J. Exper. Physiol. 1938. 28. 33.
Selladurai, S., and S. Wright, Ibidem 1932 a. 22. 233.

Selladurai, S., and S. Wright, Ibidem 1932 b. 22. 285.

Seto, H., Z. Zellforsch. u. mikr. Anat. 1935. 22. 213.

SiciLiANO, L., Arch. ital. Biol. 1900. 33. 338.

Smyth, D. H., J. Physiol. 1937. 88. 425.

Stella, G., Ibidem 1936. 87. 48 P.

Sunder-Plassmann, P., Z. Anat. u. Entw. Gesch. 1930. 93. 567.
Tschernjachiwsky, A., Trav. Labor. Rech. Biol. Univ. Madrid 1929. 26. 75.
Tello, j. F., Ibidem 1923. 21. 257.

Trinci, G., Mem. Accad. Sci. Inst. Bologna 1907. s IV.

Watzka, M., Verhandl. d. Anat. Gesellsch. 39 Vers. Amsterdam. 1930. 185.
Wiesel, j., Wien. klin. Wschr. 1906. 723.

Winder, C. V., A.m. J. Physiol. 1938. 123. 217.

Witt, D. B., L. N. Katz and L. Kohn, Ibidem 1934. 107. 213.

-Wright, S., J. Physiol. 1930. 69. 493.

VTricht, S., Quart. J. Expcr. Physiol. 1934. 24 . 169.
Wright, S., J. Pharmacol. 1935. 54. 1.

Wright. S., Quart. J. Exper. Physiol. 1937—38. 26 . 63.
ZoTTERMAN, y., Skand. Arcli, Physiol, 1935. 72 . 73.
ZoTTERMAN, Y., Ihidein 1936. 75. 105.

ZorrERMAN, Y., Acta Physiol. Scaiul. 1944. 8 . 377.
ZucKERKANDi,, E., Anat. .Anz. Erp.-II. 1901, 19 . 95.

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Acta Physiologica Scandinavica"

See other formats